CINXE.COM

<!DOCTYPE html><html class="__variable_d06c34 __variable_b20d9a kui-dark-theme bg-transparent font-sans" id="app" lang="en"><head><meta charSet="utf-8"/><meta name="viewport" content="width=device-width, initial-scale=1"/><link rel="preload" href="/_next/static/media/0064b95ffa9453c6-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/46edd4ccaf5e549e-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/4bac9051a9746ec4-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/8b83042c50f11cab-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/bb3ef058b751a6ad-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/cc0cfc6c360d4093-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/d102257ac65edb2a-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/d4aa046a6a806d8f-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="preload" href="/_next/static/media/db503b9c104834ef-s.p.woff2" as="font" crossorigin="" type="font/woff2"/><link rel="stylesheet" href="/_next/static/css/8f94c4dae2af4daf.css" data-precedence="next"/><link rel="stylesheet" href="/_next/static/css/9a3720f649c4b700.css" data-precedence="next"/><link rel="preload" as="script" fetchPriority="low" href="/_next/static/chunks/webpack-5cf9b03c4685f4fa.js"/><script src="/_next/static/chunks/c35cf5d6-6603f7fa5c2a871b.js" async=""></script><script src="/_next/static/chunks/15431-25808c8e97dc7397.js" async=""></script><script src="/_next/static/chunks/main-app-b2ef039a404a15e6.js" async=""></script><script src="/_next/static/chunks/cef99e24-4550e53c8d136ca7.js" async=""></script><script src="/_next/static/chunks/a9670928-1c89a4311971ff31.js" async=""></script><script src="/_next/static/chunks/cb9c0786-198fb9a50c9ef41d.js" async=""></script><script src="/_next/static/chunks/e6046c14-e62a3eefe5b72ddd.js" async=""></script><script src="/_next/static/chunks/21610-7deb24ce331f2ac0.js" async=""></script><script src="/_next/static/chunks/13051-735cafa2c76b6612.js" async=""></script><script src="/_next/static/chunks/67406-3f94c755e5e6329f.js" async=""></script><script src="/_next/static/chunks/40459-2ff37b28b5fc1931.js" async=""></script><script src="/_next/static/chunks/10776-f8ba0fb8bdc5c1e0.js" async=""></script><script src="/_next/static/chunks/40726-b630e3e744bb7ac9.js" async=""></script><script src="/_next/static/chunks/60401-b0b40c42264c1e12.js" async=""></script><script src="/_next/static/chunks/70594-4a7d5e3efd2703ed.js" async=""></script><script src="/_next/static/chunks/58746-6d2311b776932c61.js" async=""></script><script src="/_next/static/chunks/16226-f37a631a5aab5d92.js" async=""></script><script src="/_next/static/chunks/39013-9472e74a6e6708c2.js" async=""></script><script src="/_next/static/chunks/83635-1a0e4c2cba2f2a76.js" async=""></script><script src="/_next/static/chunks/70953-edd738fa620d4ac1.js" async=""></script><script src="/_next/static/chunks/74846-2d3d81d1f42cfeb8.js" async=""></script><script src="/_next/static/chunks/7073-d4bd33567638e78d.js" async=""></script><script src="/_next/static/chunks/7465-cc6b0469e06e7696.js" async=""></script><script src="/_next/static/chunks/92340-f455198eac6d7201.js" async=""></script><script src="/_next/static/chunks/84784-71a92c03afc120a7.js" async=""></script><script src="/_next/static/chunks/82074-df560947c012b047.js" async=""></script><script src="/_next/static/chunks/25088-cd1635fc1fffeb0c.js" async=""></script><script src="/_next/static/chunks/4518-a3e00c8ba42e3b6c.js" async=""></script><script src="/_next/static/chunks/23362-ec1a0f0c6d917ab0.js" async=""></script><script src="/_next/static/chunks/12544-fac2301ea4b12d45.js" async=""></script><script src="/_next/static/chunks/30841-02381d56be476691.js" async=""></script><script src="/_next/static/chunks/90967-e90f8a4d90679f87.js" async=""></script><script src="/_next/static/chunks/71931-dfd1072f67f502b5.js" async=""></script><script src="/_next/static/chunks/4790-d19c0f2d8ca8f6d9.js" async=""></script><script src="/_next/static/chunks/46128-e96fd11a1604d3a4.js" async=""></script><script src="/_next/static/chunks/39065-bd1cbbcb5fbe806d.js" async=""></script><script src="/_next/static/chunks/app/(app-bar-layout)/layout-e6b833b293a6a48d.js" async=""></script><script src="/_next/static/chunks/app/(app-bar-layout)/explore/layout-bfa292509e5202a5.js" async=""></script><script src="/_next/static/chunks/66697-63d4ca0711f91414.js" async=""></script><script src="/_next/static/chunks/19697-280b3bc2e9516441.js" async=""></script><script src="/_next/static/chunks/14195-243ccf9b70449304.js" async=""></script><script src="/_next/static/chunks/84327-0b3e44af5992e15b.js" async=""></script><script src="/_next/static/chunks/86916-670dd431128fa3c7.js" async=""></script><script src="/_next/static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js" async=""></script><link rel="preload" href="/_next/static/css/11122b5d7d585ae7.css" as="style"/><link rel="preload" href="/_next/static/css/fd88663f7ac2a1fd.css" as="style"/><link rel="preload" href="/_next/static/css/caf16cecf0d5eb2d.css" as="style"/><link rel="preload" href="/_next/static/css/688be17216d781f9.css" as="style"/><link rel="preload" href="/_next/static/css/b716460f5ba1aa64.css" as="style"/><link rel="preload" href="/_next/static/css/15335b524c9ed37d.css" as="style"/><link rel="preload" href="/_next/static/css/9ab03ffa4c735f8f.css" as="style"/><link rel="preload" href="//assets.adobedtm.com/5d4962a43b79/a13a52e2684e/launch-f1e5e96c44d5.min.js" as="script"/><link rel="preload" href="https://cdn.cookielaw.org/scripttemplates/otSDKStub.js" as="script"/><link rel="preload" href="https://images.nvidia.com/aem-dam/Solutions/ot-js/ot-custom.js" as="script"/><title>Try NVIDIA NIM APIs</title><meta name="description" content="Experience the leading models to build enterprise generative AI apps now."/><link rel="canonical" href="https://build.nvidia.com"/><meta property="og:title" content="Try NVIDIA NIM APIs"/><meta property="og:description" content="Experience the leading models to build enterprise generative AI apps now."/><meta property="og:image:type" content="image/jpeg"/><meta property="og:image:width" content="1200"/><meta property="og:image:height" content="630"/><meta property="og:image" content="https://build.nvidia.com/opengraph-image.jpg?6ec102a0470b935b"/><meta property="og:type" content="website"/><meta name="twitter:card" content="summary_large_image"/><meta name="twitter:title" content="Try NVIDIA NIM APIs"/><meta name="twitter:description" content="Experience the leading models to build enterprise generative AI apps now."/><meta name="twitter:image:type" content="image/jpeg"/><meta name="twitter:image:width" content="1200"/><meta name="twitter:image:height" content="630"/><meta name="twitter:image" content="https://build.nvidia.com/twitter-image.jpg?6ec102a0470b935b"/><link rel="icon" href="/favicon.ico" type="image/x-icon" sizes="48x48"/><meta name="next-size-adjust"/><script src="/_next/static/chunks/polyfills-42372ed130431b0a.js" noModule=""></script><style id="stitches">--sxs{--sxs:0 t-jeUjvk t-laoRNc t-gNphrk}@media{:root,.t-jeUjvk{--borderRadius-xl:16px;--borderRadius-lg:8px;--borderRadius-md:4px;--borderRadius-sm:2px;--borderRadius-sharp:0px;--borderWidths-lg:4px;--borderWidths-md:2px;--borderWidths-sm:1px;--colors-brand:#76b900;--colors-n000:#ffffff;--colors-n025:#fafafa;--colors-n050:#f2f2f2;--colors-n100:#e2e2e2;--colors-n200:#d8d8d8;--colors-n300:#c6c6c6;--colors-n400:#8f8f8f;--colors-n500:#767676;--colors-n600:#5e5e5e;--colors-n700:#343434;--colors-n800:#292929;--colors-n900:#202020;--colors-n950:#121212;--colors-n1000:#000000;--colors-g050:#ecfad9;--colors-g100:#d6f2ad;--colors-g200:#bfe882;--colors-g300:#a7db57;--colors-g400:#8fcb2b;--colors-g500:#76b900;--colors-g600:#57920d;--colors-g700:#407015;--colors-g800:#2f521a;--colors-g900:#23371a;--colors-g950:#192116;--colors-b050:#e9f4fb;--colors-b100:#cde6fa;--colors-b200:#badef8;--colors-b300:#9accf2;--colors-b400:#3a96d9;--colors-b500:#2378ca;--colors-b600:#2a63ba;--colors-b700:#013076;--colors-b800:#00265e;--colors-b900:#001e4b;--colors-b950:#00112c;--colors-y050:#f9f4ca;--colors-y100:#f6e383;--colors-y200:#f3d85d;--colors-y300:#e7c32b;--colors-y400:#c58420;--colors-y500:#a96700;--colors-y600:#924c00;--colors-y700:#5a2400;--colors-y800:#491c00;--colors-y900:#3a1500;--colors-y950:#220c00;--colors-r050:#ffefed;--colors-r100:#ffdad6;--colors-r200:#ffceca;--colors-r300:#feb3ad;--colors-r400:#ec6358;--colors-r500:#dc3528;--colors-r600:#b6251b;--colors-r700:#6c0d08;--colors-r800:#570704;--colors-r900:#480200;--colors-r950:#2d0100;--colors-t050:#e1f7f3;--colors-t100:#b7eee4;--colors-t200:#a0e6d9;--colors-t300:#7cd7c7;--colors-t400:#1fa18d;--colors-t500:#008471;--colors-t600:#006b5e;--colors-t700:#003b35;--colors-t800:#002f2b;--colors-t900:#002522;--colors-t950:#001514;--colors-p050:#fbeefe;--colors-p100:#f1dcf7;--colors-p200:#ead1f3;--colors-p300:#debae8;--colors-p400:#b579c6;--colors-p500:#a05ab4;--colors-p600:#8c3ca0;--colors-p700:#5a066d;--colors-p800:#490457;--colors-p900:#3b0047;--colors-p950:#25002c;--colors-tk000:rgba(0,0,0,0.00);--colors-tk050:rgba(0,0,0,0.05);--colors-tk100:rgba(0,0,0,0.10);--colors-tk120:rgba(0,0,0,0.12);--colors-tk150:rgba(0,0,0,0.15);--colors-tk200:rgba(0,0,0,0.20);--colors-tk300:rgba(0,0,0,0.30);--colors-tk400:rgba(0,0,0,0.40);--colors-tk500:rgba(0,0,0,0.50);--colors-tk600:rgba(0,0,0,0.60);--colors-tk700:rgba(0,0,0,0.70);--colors-tk800:rgba(0,0,0,0.80);--colors-tk900:rgba(0,0,0,0.90);--colors-tw000:rgba(255,255,255,0.00);--colors-tw050:rgba(255,255,255,0.05);--colors-tw100:rgba(255,255,255,0.10);--colors-tw120:rgba(255,255,255,0.12);--colors-tw250:rgba(255,255,255,0.25);--colors-tw200:rgba(255,255,255,0.20);--colors-tw300:rgba(255,255,255,0.30);--colors-tw400:rgba(255,255,255,0.40);--colors-tw500:rgba(255,255,255,0.50);--colors-tw600:rgba(255,255,255,0.60);--colors-tw700:rgba(255,255,255,0.70);--colors-tw800:rgba(255,255,255,0.80);--colors-tw900:rgba(255,255,255,0.90);--colors-displayBgHighest:#ffffff;--colors-displayBgHighestHover:#f2f2f2;--colors-displayBgHighestActive:#e2e2e2;--colors-displayBgHigher:#ffffff;--colors-displayBgHigherHover:#f2f2f2;--colors-displayBgHigherActive:#e2e2e2;--colors-displayBgHigh:#ffffff;--colors-displayBgHighHover:#f2f2f2;--colors-displayBgHighActive:#e2e2e2;--colors-displayBgMid:#ffffff;--colors-displayBgMidHover:#f2f2f2;--colors-displayBgMidActive:#e2e2e2;--colors-displayBgLow:#f2f2f2;--colors-displayBgLowHover:#e2e2e2;--colors-displayBgLowActive:#d8d8d8;--colors-displayBgCanvas:#fafafa;--colors-displayBgAccent1Hover:#e9f4fb;--colors-displayBgAccent1Active:#2a63ba;--colors-displayBgAccent1Solid:#e9f4fb;--colors-displayBgAccent1SolidHover:#cde6fa;--colors-displayBgAccent1SolidActive:#2a63ba;--colors-displayBgAccent1Bold:#2a63ba;--colors-displayBgAccent1Disabled:#8f8f8f;--colors-displayBgAccent2Hover:#ecfad9;--colors-displayBgAccent2Active:#57920d;--colors-displayBgAccent2Solid:#ecfad9;--colors-displayBgAccent2SolidHover:#d6f2ad;--colors-displayBgAccent2SolidActive:#57920d;--colors-displayBgAccent2Bold:#57920d;--colors-displayBgAccent2Disabled:#8f8f8f;--colors-displayBgAccent3Hover:#f9f4ca;--colors-displayBgAccent3Active:#924c00;--colors-displayBgAccent3Solid:#f9f4ca;--colors-displayBgAccent3SolidHover:#f6e383;--colors-displayBgAccent3SolidActive:#924c00;--colors-displayBgAccent3Bold:#924c00;--colors-displayBgAccent3Disabled:#8f8f8f;--colors-displayBgAccent4Hover:#ffefed;--colors-displayBgAccent4Active:#b6251b;--colors-displayBgAccent4Solid:#ffefed;--colors-displayBgAccent4SolidHover:#ffdad6;--colors-displayBgAccent4SolidActive:#b6251b;--colors-displayBgAccent4Bold:#b6251b;--colors-displayBgAccent4Disabled:#8f8f8f;--colors-displayBgAccent5Hover:#fbeefe;--colors-displayBgAccent5Active:#8c3ca0;--colors-displayBgAccent5Solid:#fbeefe;--colors-displayBgAccent5SolidHover:#f1dcf7;--colors-displayBgAccent5SolidActive:#8c3ca0;--colors-displayBgAccent5Bold:#8c3ca0;--colors-displayBgAccent5Disabled:#8f8f8f;--colors-displayBgAccent6Hover:#e1f7f3;--colors-displayBgAccent6Active:#006b5e;--colors-displayBgAccent6Solid:#e1f7f3;--colors-displayBgAccent6SolidHover:#b7eee4;--colors-displayBgAccent6SolidActive:#006b5e;--colors-displayBgAccent6Bold:#006b5e;--colors-displayBgAccent6Disabled:#8f8f8f;--colors-displayBgAccent7Hover:#e2e2e2;--colors-displayBgAccent7Active:#343434;--colors-displayBgAccent7Solid:#d8d8d8;--colors-displayBgAccent7SolidHover:#c6c6c6;--colors-displayBgAccent7SolidActive:#343434;--colors-displayBgAccent7Bold:#343434;--colors-displayBgAccent7Disabled:#8f8f8f;--colors-displayFgHighestPrimary:#202020;--colors-displayFgHighestSecondary:#5e5e5e;--colors-displayFgHighestDisabled:#767676;--colors-displayFgHighestPlaceholder:#8f8f8f;--colors-displayFgHighestStrong:#000000;--colors-displayFgHighestSelected:#76b900;--colors-displayFgHighestSelectedHover:#57920d;--colors-displayFgHigherPrimary:#202020;--colors-displayFgHigherSecondary:#5e5e5e;--colors-displayFgHigherDisabled:#767676;--colors-displayFgHigherPlaceholder:#8f8f8f;--colors-displayFgHigherStrong:#000000;--colors-displayFgHigherSelected:#76b900;--colors-displayFgHigherSelectedHover:#57920d;--colors-displayFgHighPrimary:#202020;--colors-displayFgHighSecondary:#5e5e5e;--colors-displayFgHighDisabled:#767676;--colors-displayFgHighPlaceholder:#8f8f8f;--colors-displayFgHighStrong:#000000;--colors-displayFgHighSelected:#76b900;--colors-displayFgHighSelectedHover:#57920d;--colors-displayFgMidPrimary:#202020;--colors-displayFgMidSecondary:#5e5e5e;--colors-displayFgMidDisabled:#767676;--colors-displayFgMidPlaceholder:#8f8f8f;--colors-displayFgMidStrong:#000000;--colors-displayFgMidSelected:#76b900;--colors-displayFgMidSelectedHover:#57920d;--colors-displayFgLowPrimary:#202020;--colors-displayFgLowSecondary:#5e5e5e;--colors-displayFgLowDisabled:#767676;--colors-displayFgLowPlaceholder:#8f8f8f;--colors-displayFgLowStrong:#000000;--colors-displayFgLowSelected:#76b900;--colors-displayFgLowSelectedHover:#57920d;--colors-displayFgAccent1:#2a63ba;--colors-displayFgAccent1Solid:#2a63ba;--colors-displayFgAccent1Bold:#e9f4fb;--colors-displayFgAccent1Disabled:#8f8f8f;--colors-displayFgAccent2:#57920d;--colors-displayFgAccent2Solid:#57920d;--colors-displayFgAccent2Bold:#ecfad9;--colors-displayFgAccent2Disabled:#8f8f8f;--colors-displayFgAccent3:#924c00;--colors-displayFgAccent3Solid:#924c00;--colors-displayFgAccent3Bold:#f9f4ca;--colors-displayFgAccent3Disabled:#8f8f8f;--colors-displayFgAccent4:#b6251b;--colors-displayFgAccent4Solid:#b6251b;--colors-displayFgAccent4Bold:#ffefed;--colors-displayFgAccent4Disabled:#8f8f8f;--colors-displayFgAccent5:#8c3ca0;--colors-displayFgAccent5Solid:#8c3ca0;--colors-displayFgAccent5Bold:#fbeefe;--colors-displayFgAccent5Disabled:#8f8f8f;--colors-displayFgAccent6:#006b5e;--colors-displayFgAccent6Solid:#006b5e;--colors-displayFgAccent6Bold:#e1f7f3;--colors-displayFgAccent6Disabled:#8f8f8f;--colors-displayFgAccent7:#343434;--colors-displayFgAccent7Solid:#343434;--colors-displayFgAccent7Bold:#f2f2f2;--colors-displayFgAccent7Disabled:#8f8f8f;--colors-displayFgInverse:#ffffff;--colors-displayFgSubtle:#f2f2f2;--colors-displayBorderHighest:#d8d8d8;--colors-displayBorderHighestHover:#8f8f8f;--colors-displayBorderHighestActive:#5e5e5e;--colors-displayBorderHighestDisabled:#e2e2e2;--colors-displayBorderHighestBase:#ffffff;--colors-displayBorderHighestSelected:#76b900;--colors-displayBorderHighestSelectedHover:#57920d;--colors-displayBorderHigher:#d8d8d8;--colors-displayBorderHigherHover:#8f8f8f;--colors-displayBorderHigherActive:#5e5e5e;--colors-displayBorderHigherDisabled:#e2e2e2;--colors-displayBorderHigherBase:#ffffff;--colors-displayBorderHigherSelected:#76b900;--colors-displayBorderHigherSelectedHover:#57920d;--colors-displayBorderHigh:#d8d8d8;--colors-displayBorderHighHover:#8f8f8f;--colors-displayBorderHighActive:#5e5e5e;--colors-displayBorderHighDisabled:#e2e2e2;--colors-displayBorderHighBase:#ffffff;--colors-displayBorderHighSelected:#76b900;--colors-displayBorderHighSelectedHover:#57920d;--colors-displayBorderMid:#d8d8d8;--colors-displayBorderMidHover:#8f8f8f;--colors-displayBorderMidActive:#5e5e5e;--colors-displayBorderMidDisabled:#e2e2e2;--colors-displayBorderMidBase:#ffffff;--colors-displayBorderMidSelected:#76b900;--colors-displayBorderMidSelectedHover:#57920d;--colors-displayBorderLow:#d8d8d8;--colors-displayBorderLowHover:#8f8f8f;--colors-displayBorderLowActive:#5e5e5e;--colors-displayBorderLowDisabled:#e2e2e2;--colors-displayBorderLowBase:#f2f2f2;--colors-displayBorderLowSelected:#76b900;--colors-displayBorderLowSelectedHover:#57920d;--colors-displayBorderAccent1:#2a63ba;--colors-displayBorderAccent1Solid:#e9f4fb;--colors-displayBorderAccent1SolidHover:#cde6fa;--colors-displayBorderAccent1SolidActive:#2a63ba;--colors-displayBorderAccent1Bold:#2a63ba;--colors-displayBorderAccent1Disabled:#8f8f8f;--colors-displayBorderAccent2:#57920d;--colors-displayBorderAccent2Solid:#ecfad9;--colors-displayBorderAccent2SolidHover:#d6f2ad;--colors-displayBorderAccent2SolidActive:#57920d;--colors-displayBorderAccent2Bold:#57920d;--colors-displayBorderAccent2Disabled:#8f8f8f;--colors-displayBorderAccent3:#924c00;--colors-displayBorderAccent3Solid:#f9f4ca;--colors-displayBorderAccent3SolidHover:#f6e383;--colors-displayBorderAccent3SolidActive:#924c00;--colors-displayBorderAccent3Bold:#924c00;--colors-displayBorderAccent3Disabled:#8f8f8f;--colors-displayBorderAccent4:#b6251b;--colors-displayBorderAccent4Solid:#ffefed;--colors-displayBorderAccent4SolidHover:#ffdad6;--colors-displayBorderAccent4SolidActive:#b6251b;--colors-displayBorderAccent4Bold:#b6251b;--colors-displayBorderAccent4Disabled:#8f8f8f;--colors-displayBorderAccent5:#8c3ca0;--colors-displayBorderAccent5Solid:#fbeefe;--colors-displayBorderAccent5SolidHover:#f1dcf7;--colors-displayBorderAccent5SolidActive:#8c3ca0;--colors-displayBorderAccent5Bold:#8c3ca0;--colors-displayBorderAccent5Disabled:#8f8f8f;--colors-displayBorderAccent6:#006b5e;--colors-displayBorderAccent6Solid:#e1f7f3;--colors-displayBorderAccent6SolidHover:#b7eee4;--colors-displayBorderAccent6SolidActive:#006b5e;--colors-displayBorderAccent6Bold:#006b5e;--colors-displayBorderAccent6Disabled:#8f8f8f;--colors-displayBorderAccent7:#343434;--colors-displayBorderAccent7Solid:#d8d8d8;--colors-displayBorderAccent7SolidHover:#c6c6c6;--colors-displayBorderAccent7SolidActive:#343434;--colors-displayBorderAccent7Bold:#343434;--colors-displayBorderAccent7Disabled:#8f8f8f;--colors-displayBorderStatic:#c6c6c6;--colors-menuBgDefault:#ffffff;--colors-menuBgHover:#f2f2f2;--colors-menuBgActive:#e2e2e2;--colors-menuBgComplete:#8fcb2b;--colors-menuBgSelected:#f2f2f2;--colors-menuBgDisabled:#ffffff;--colors-menuBgPlaceholder:#ffffff;--colors-menuBgDanger:#ffffff;--colors-menuBgDangerHover:#ffefed;--colors-menuBgBase:#ffffff;--colors-menuBgBaseSelected:#ffffff;--colors-menuFgDefault:#5e5e5e;--colors-menuFgHover:#202020;--colors-menuFgActive:#202020;--colors-menuFgComplete:#202020;--colors-menuFgSelected:#202020;--colors-menuFgDisabled:#767676;--colors-menuFgPlaceholder:#8f8f8f;--colors-menuFgDanger:#dc3528;--colors-menuFgDangerHover:#dc3528;--colors-menuFgPrimary:#202020;--colors-menuFgHeader:#5e5e5e;--colors-menuFgFooter:#5e5e5e;--colors-menuBorderDefault:#d8d8d8;--colors-menuBorderHover:#8f8f8f;--colors-menuBorderActive:#d8d8d8;--colors-menuBorderComplete:#76b900;--colors-menuBorderSelected:#76b900;--colors-menuBorderDisabled:#d8d8d8;--colors-menuBorderPlaceholder:#d8d8d8;--colors-inputBgDefault:#ffffff;--colors-inputBgHover:#ffffff;--colors-inputBgActive:#f2f2f2;--colors-inputBgComplete:#ffffff;--colors-inputBgDisabled:#f2f2f2;--colors-inputBgError:#f2f2f2;--colors-inputBgSuccess:#ffffff;--colors-inputFgDefault:#8f8f8f;--colors-inputFgHover:#8f8f8f;--colors-inputFgActive:#202020;--colors-inputFgComplete:#202020;--colors-inputFgDisabled:#767676;--colors-inputFgError:#202020;--colors-inputFgErrorMessage:#dc3528;--colors-inputFgSuccess:#202020;--colors-inputFgSuccessMessage:#76b900;--colors-inputFgLabel:#5e5e5e;--colors-inputFgRequired:#dc3528;--colors-inputFgIcon:#5e5e5e;--colors-inputFgIconHover:#202020;--colors-inputFgIconActive:#202020;--colors-inputFgIconComplete:#5e5e5e;--colors-inputFgIconDisabled:#767676;--colors-inputFgIconError:#dc3528;--colors-inputFgIconSuccess:#76b900;--colors-inputBorderDefault:#c6c6c6;--colors-inputBorderHover:#767676;--colors-inputBorderActive:#76b900;--colors-inputBorderComplete:#c6c6c6;--colors-inputBorderDisabled:#d8d8d8;--colors-inputBorderError:#dc3528;--colors-inputBorderSuccess:#76b900;--colors-actionBgDefault:#ffffff;--colors-actionBgHover:#e2e2e2;--colors-actionBgActive:#c6c6c6;--colors-actionBgDisabled:#f2f2f2;--colors-actionBgSelected:#ecfad9;--colors-actionBgSelectedHover:#d6f2ad;--colors-actionBgSelectedActive:#a7db57;--colors-actionBgSelectedDisabled:#f2f2f2;--colors-actionBgError:#ffefed;--colors-actionBgErrorHover:#ffdad6;--colors-actionBgErrorActive:#feb3ad;--colors-actionBgErrorDisabled:#f2f2f2;--colors-actionBgPrimary:#76b900;--colors-actionBgPrimaryHover:#8fcb2b;--colors-actionBgPrimaryActive:#a7db57;--colors-actionBgPrimaryDisabled:#d8d8d8;--colors-actionBgSecondary:rgba(0,0,0,0.00);--colors-actionBgSecondaryHover:rgba(0,0,0,0.05);--colors-actionBgSecondaryActive:rgba(0,0,0,0.15);--colors-actionBgSecondaryDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiary:rgba(0,0,0,0.00);--colors-actionBgTertiaryHover:rgba(0,0,0,0.05);--colors-actionBgTertiaryActive:rgba(0,0,0,0.15);--colors-actionBgTertiaryDisabled:rgba(0,0,0,0.00);--colors-actionBgSecondaryInverse:rgba(0,0,0,0.00);--colors-actionBgSecondaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgSecondaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgSecondaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverse:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgTertiaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgTertiaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgPrimaryDanger:#dc3528;--colors-actionBgPrimaryDangerHover:#b6251b;--colors-actionBgPrimaryDangerActive:#570704;--colors-actionBgPrimaryDangerDisabled:#d8d8d8;--colors-actionBgSecondaryDanger:rgba(0,0,0,0.00);--colors-actionBgSecondaryDangerHover:#ffdad6;--colors-actionBgSecondaryDangerActive:#feb3ad;--colors-actionBgSecondaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryDanger:rgba(0,0,0,0.00);--colors-actionBgTertiaryDangerHover:#ffdad6;--colors-actionBgTertiaryDangerActive:#feb3ad;--colors-actionBgTertiaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgBase:#c6c6c6;--colors-actionBgBaseDisabled:#e2e2e2;--colors-actionBgOn:#76b900;--colors-actionBgOnDisabled:#a7db57;--colors-actionFgDefault:#202020;--colors-actionFgHover:#202020;--colors-actionFgActive:#202020;--colors-actionFgDisabled:#767676;--colors-actionFgStrong:#000000;--colors-actionFgSelected:#76b900;--colors-actionFgSelectedHover:#57920d;--colors-actionFgSelectedActive:#407015;--colors-actionFgSelectedDisabled:#767676;--colors-actionFgError:#dc3528;--colors-actionFgErrorHover:#b6251b;--colors-actionFgErrorActive:#570704;--colors-actionFgErrorDisabled:#767676;--colors-actionFgPrimary:#202020;--colors-actionFgPrimaryHover:#202020;--colors-actionFgPrimaryActive:#202020;--colors-actionFgPrimaryDisabled:#767676;--colors-actionFgSecondary:#202020;--colors-actionFgSecondaryHover:#202020;--colors-actionFgSecondaryActive:#202020;--colors-actionFgSecondaryDisabled:#8f8f8f;--colors-actionFgTertiary:#202020;--colors-actionFgTertiaryHover:#202020;--colors-actionFgTertiaryActive:#202020;--colors-actionFgTertiaryDisabled:#8f8f8f;--colors-actionFgSecondaryInverse:#ffffff;--colors-actionFgSecondaryInverseHover:#ffffff;--colors-actionFgSecondaryInverseActive:#ffffff;--colors-actionFgSecondaryInverseDisabled:#d8d8d8;--colors-actionFgTertiaryInverse:#ffffff;--colors-actionFgTertiaryInverseHover:#ffffff;--colors-actionFgTertiaryInverseActive:#ffffff;--colors-actionFgTertiaryInverseDisabled:#d8d8d8;--colors-actionFgPrimaryDanger:#ffffff;--colors-actionFgPrimaryDangerHover:#ffffff;--colors-actionFgPrimaryDangerActive:#ffffff;--colors-actionFgPrimaryDangerDisabled:#767676;--colors-actionFgSecondaryDanger:#dc3528;--colors-actionFgSecondaryDangerHover:#b6251b;--colors-actionFgSecondaryDangerActive:#6c0d08;--colors-actionFgSecondaryDangerDisabled:#8f8f8f;--colors-actionFgTertiaryDanger:#dc3528;--colors-actionFgTertiaryDangerHover:#b6251b;--colors-actionFgTertiaryDangerActive:#6c0d08;--colors-actionFgTertiaryDangerDisabled:#8f8f8f;--colors-actionFgInverse:#ffffff;--colors-actionFgOn:#ffffff;--colors-actionFgOnDisabled:#f2f2f2;--colors-actionBorderDefault:#8f8f8f;--colors-actionBorderHover:#767676;--colors-actionBorderActive:#343434;--colors-actionBorderDisabled:#c6c6c6;--colors-actionBorderSelected:#76b900;--colors-actionBorderSelectedHover:#57920d;--colors-actionBorderSelectedActive:#407015;--colors-actionBorderSelectedDisabled:#d8d8d8;--colors-actionBorderError:#dc3528;--colors-actionBorderErrorHover:#b6251b;--colors-actionBorderErrorActive:#570704;--colors-actionBorderErrorDisabled:#d8d8d8;--colors-actionBorderPrimary:#76b900;--colors-actionBorderPrimaryHover:#a7db57;--colors-actionBorderPrimaryActive:#a7db57;--colors-actionBorderPrimaryDisabled:#d8d8d8;--colors-actionBorderSecondary:#5e5e5e;--colors-actionBorderSecondaryHover:#202020;--colors-actionBorderSecondaryActive:#202020;--colors-actionBorderSecondaryDisabled:#8f8f8f;--colors-actionBorderSecondaryInverse:#ffffff;--colors-actionBorderSecondaryInverseHover:#ffffff;--colors-actionBorderSecondaryInverseActive:#ffffff;--colors-actionBorderSecondaryInverseDisabled:#d8d8d8;--colors-actionBorderPrimaryDanger:#dc3528;--colors-actionBorderPrimaryDangerHover:#570704;--colors-actionBorderPrimaryDangerActive:#570704;--colors-actionBorderPrimaryDangerDisabled:#d8d8d8;--colors-actionBorderSecondaryDanger:#dc3528;--colors-actionBorderSecondaryDangerHover:#dc3528;--colors-actionBorderSecondaryDangerActive:#6c0d08;--colors-actionBorderSecondaryDangerDisabled:#8f8f8f;--colors-actionBorderInverse:#ffffff;--colors-actionBorderBase:#8f8f8f;--colors-feedbackBgInfo:#e9f4fb;--colors-feedbackBgInfoSolid:#2378ca;--colors-feedbackBgSuccess:#ecfad9;--colors-feedbackBgSuccessSolid:#76b900;--colors-feedbackBgCaution:#f9f4ca;--colors-feedbackBgCautionSolid:#e7c32b;--colors-feedbackBgError:#ffefed;--colors-feedbackBgErrorSolid:#dc3528;--colors-feedbackBgNeutral:#e2e2e2;--colors-feedbackBgProgress:#ffffff;--colors-feedbackBgTooltip:#343434;--colors-feedbackBgBase:#ffffff;--colors-feedbackFgInfo:#2378ca;--colors-feedbackFgInfoHover:#2a63ba;--colors-feedbackFgInfoActive:#013076;--colors-feedbackFgSuccess:#76b900;--colors-feedbackFgSuccessHover:#57920d;--colors-feedbackFgSuccessActive:#407015;--colors-feedbackFgCaution:#a96700;--colors-feedbackFgCautionHover:#924c00;--colors-feedbackFgCautionActive:#5a2400;--colors-feedbackFgError:#dc3528;--colors-feedbackFgErrorHover:#b6251b;--colors-feedbackFgErrorActive:#6c0d08;--colors-feedbackFgNeutral:#5e5e5e;--colors-feedbackFgNeutralHover:#343434;--colors-feedbackFgNeutralActive:#000000;--colors-feedbackFgProgress:#76b900;--colors-feedbackFgProgressActive:#407015;--colors-feedbackFgTooltip:#f2f2f2;--colors-feedbackFgPrimary:#202020;--colors-feedbackFgSecondary:#5e5e5e;--colors-feedbackFgInverse:#ffffff;--colors-feedbackFgReverse:#202020;--colors-feedbackBorderInfo:#2378ca;--colors-feedbackBorderSuccess:#76b900;--colors-feedbackBorderCaution:#a96700;--colors-feedbackBorderError:#dc3528;--colors-feedbackBorderNeutral:#8f8f8f;--colors-feedbackBorderProgress:#d8d8d8;--colors-feedbackBorderTooltip:#d8d8d8;--fonts-nvidiaSans:"NVIDIA Sans", "NVIDIA Sans Fallback";--fonts-jetbrainsMono:"JetBrains Mono", "JetBrains Mono Fallback";--fontSizes-xxl:48px;--fontSizes-xl:32px;--fontSizes-lg:24px;--fontSizes-ml:20px;--fontSizes-md:16px;--fontSizes-ms:14px;--fontSizes-sm:12px;--fontSizes-xs:10px;--fontWeights-light:300;--fontWeights-regular:400;--fontWeights-medium:500;--fontWeights-bold:700;--letterSpacings-less:-0.015rem;--letterSpacings-none:0rem;--letterSpacings-more:0.05rem;--lineHeights-caps:90%;--lineHeights-text:100%;--lineHeights-heading:120%;--lineHeights-body:150%;--shadows-lightHigher:0px 8px 10px 0 #00000026;--shadows-lightHigh:0px 4px 5px 0 #0000001f;--shadows-lightMid:0px 2px 3px 0 #00000024;--shadows-lightLow:0px 1px 3px 0 #00000033;--shadows-darkHigher:0px 8px 10px 0 #0000004d;--shadows-darkHigh:0px 4px 5px 0 #0000003d;--shadows-darkMid:0px 2px 4px 0 #00000047;--shadows-darkLow:0px 1px 3px 0 #00000066;--shadows-displayModal:0px 8px 10px 0 #00000026;--shadows-displayPageHeader:0px 2px 3px 0 #00000024;--shadows-displayPopover:0px 4px 5px 0 #0000001f;--shadows-displayCarouselButton:0px 2px 3px 0 #00000024;--shadows-displaySidePanelRight:-4px 0px 5px 0 #0000001f;--shadows-displaySidePanelLeft:4px 0px 5px 0 #0000001f;--shadows-feedbackTooltip:0px 4px 5px 0 #0000001f;--shadows-feedbackNotification:0px 4px 5px 0 #0000001f;--shadows-menuDropdown:0px 4px 5px 0 #0000001f;--shadows-menuPopout:0px 2px 3px 0 #00000024;--shadows-inputToggle:0px 1px 3px 0 #00000033;--space-xxl:64px;--space-xl:48px;--space-lg:32px;--space-ml:24px;--space-md:16px;--space-ms:12px;--space-sm:8px;--space-ss:6px;--space-xs:4px;--space-xxs:2px;--space-none:0px}.t-laoRNc{--colors-brand:#76b900;--colors-n000:#ffffff;--colors-n025:#fafafa;--colors-n050:#f2f2f2;--colors-n100:#e2e2e2;--colors-n200:#d8d8d8;--colors-n300:#c6c6c6;--colors-n400:#8f8f8f;--colors-n500:#767676;--colors-n600:#5e5e5e;--colors-n700:#343434;--colors-n800:#292929;--colors-n900:#202020;--colors-n950:#121212;--colors-n1000:#000000;--colors-g050:#ecfad9;--colors-g100:#d6f2ad;--colors-g200:#bfe882;--colors-g300:#a7db57;--colors-g400:#8fcb2b;--colors-g500:#76b900;--colors-g600:#57920d;--colors-g700:#407015;--colors-g800:#2f521a;--colors-g900:#23371a;--colors-g950:#192116;--colors-b050:#e9f4fb;--colors-b100:#cde6fa;--colors-b200:#badef8;--colors-b300:#9accf2;--colors-b400:#3a96d9;--colors-b500:#2378ca;--colors-b600:#2a63ba;--colors-b700:#013076;--colors-b800:#00265e;--colors-b900:#001e4b;--colors-b950:#00112c;--colors-y050:#f9f4ca;--colors-y100:#f6e383;--colors-y200:#f3d85d;--colors-y300:#e7c32b;--colors-y400:#c58420;--colors-y500:#a96700;--colors-y600:#924c00;--colors-y700:#5a2400;--colors-y800:#491c00;--colors-y900:#3a1500;--colors-y950:#220c00;--colors-r050:#ffefed;--colors-r100:#ffdad6;--colors-r200:#ffceca;--colors-r300:#feb3ad;--colors-r400:#ec6358;--colors-r500:#dc3528;--colors-r600:#b6251b;--colors-r700:#6c0d08;--colors-r800:#570704;--colors-r900:#480200;--colors-r950:#2d0100;--colors-t050:#e1f7f3;--colors-t100:#b7eee4;--colors-t200:#a0e6d9;--colors-t300:#7cd7c7;--colors-t400:#1fa18d;--colors-t500:#008471;--colors-t600:#006b5e;--colors-t700:#003b35;--colors-t800:#002f2b;--colors-t900:#002522;--colors-t950:#001514;--colors-p050:#fbeefe;--colors-p100:#f1dcf7;--colors-p200:#ead1f3;--colors-p300:#debae8;--colors-p400:#b579c6;--colors-p500:#a05ab4;--colors-p600:#8c3ca0;--colors-p700:#5a066d;--colors-p800:#490457;--colors-p900:#3b0047;--colors-p950:#25002c;--colors-tk000:rgba(0,0,0,0.00);--colors-tk050:rgba(0,0,0,0.05);--colors-tk100:rgba(0,0,0,0.10);--colors-tk120:rgba(0,0,0,0.12);--colors-tk150:rgba(0,0,0,0.15);--colors-tk200:rgba(0,0,0,0.20);--colors-tk300:rgba(0,0,0,0.30);--colors-tk400:rgba(0,0,0,0.40);--colors-tk500:rgba(0,0,0,0.50);--colors-tk600:rgba(0,0,0,0.60);--colors-tk700:rgba(0,0,0,0.70);--colors-tk800:rgba(0,0,0,0.80);--colors-tk900:rgba(0,0,0,0.90);--colors-tw000:rgba(255,255,255,0.00);--colors-tw050:rgba(255,255,255,0.05);--colors-tw100:rgba(255,255,255,0.10);--colors-tw120:rgba(255,255,255,0.12);--colors-tw250:rgba(255,255,255,0.25);--colors-tw200:rgba(255,255,255,0.20);--colors-tw300:rgba(255,255,255,0.30);--colors-tw400:rgba(255,255,255,0.40);--colors-tw500:rgba(255,255,255,0.50);--colors-tw600:rgba(255,255,255,0.60);--colors-tw700:rgba(255,255,255,0.70);--colors-tw800:rgba(255,255,255,0.80);--colors-tw900:rgba(255,255,255,0.90);--colors-displayBgHighest:#000000;--colors-displayBgHighestHover:#202020;--colors-displayBgHighestActive:#343434;--colors-displayBgHigher:#343434;--colors-displayBgHigherHover:#292929;--colors-displayBgHigherActive:#121212;--colors-displayBgHigh:#292929;--colors-displayBgHighHover:#202020;--colors-displayBgHighActive:#000000;--colors-displayBgMid:#202020;--colors-displayBgMidHover:#343434;--colors-displayBgMidActive:#5e5e5e;--colors-displayBgLow:#121212;--colors-displayBgLowHover:#292929;--colors-displayBgLowActive:#343434;--colors-displayBgCanvas:#202020;--colors-displayBgAccent1Hover:#013076;--colors-displayBgAccent1Active:#9accf2;--colors-displayBgAccent1Solid:#9accf2;--colors-displayBgAccent1SolidHover:#badef8;--colors-displayBgAccent1SolidActive:#2a63ba;--colors-displayBgAccent1Bold:#2a63ba;--colors-displayBgAccent1Disabled:#8f8f8f;--colors-displayBgAccent2Hover:#407015;--colors-displayBgAccent2Active:#a7db57;--colors-displayBgAccent2Solid:#a7db57;--colors-displayBgAccent2SolidHover:#bfe882;--colors-displayBgAccent2SolidActive:#57920d;--colors-displayBgAccent2Bold:#57920d;--colors-displayBgAccent2Disabled:#8f8f8f;--colors-displayBgAccent3Hover:#5a2400;--colors-displayBgAccent3Active:#e7c32b;--colors-displayBgAccent3Solid:#e7c32b;--colors-displayBgAccent3SolidHover:#f3d85d;--colors-displayBgAccent3SolidActive:#924c00;--colors-displayBgAccent3Bold:#924c00;--colors-displayBgAccent3Disabled:#8f8f8f;--colors-displayBgAccent4Hover:#6c0d08;--colors-displayBgAccent4Active:#feb3ad;--colors-displayBgAccent4Solid:#feb3ad;--colors-displayBgAccent4SolidHover:#ffceca;--colors-displayBgAccent4SolidActive:#b6251b;--colors-displayBgAccent4Bold:#b6251b;--colors-displayBgAccent4Disabled:#8f8f8f;--colors-displayBgAccent5Hover:#5a066d;--colors-displayBgAccent5Active:#debae8;--colors-displayBgAccent5Solid:#debae8;--colors-displayBgAccent5SolidHover:#ead1f3;--colors-displayBgAccent5SolidActive:#8c3ca0;--colors-displayBgAccent5Bold:#8c3ca0;--colors-displayBgAccent5Disabled:#8f8f8f;--colors-displayBgAccent6Hover:#003b35;--colors-displayBgAccent6Active:#7cd7c7;--colors-displayBgAccent6Solid:#7cd7c7;--colors-displayBgAccent6SolidHover:#a0e6d9;--colors-displayBgAccent6SolidActive:#006b5e;--colors-displayBgAccent6Bold:#006b5e;--colors-displayBgAccent6Disabled:#8f8f8f;--colors-displayBgAccent7Hover:#343434;--colors-displayBgAccent7Active:#d8d8d8;--colors-displayBgAccent7Solid:#c6c6c6;--colors-displayBgAccent7SolidHover:#d8d8d8;--colors-displayBgAccent7SolidActive:#5e5e5e;--colors-displayBgAccent7Bold:#5e5e5e;--colors-displayBgAccent7Disabled:#8f8f8f;--colors-displayFgHighestPrimary:#f2f2f2;--colors-displayFgHighestSecondary:#c6c6c6;--colors-displayFgHighestDisabled:#8f8f8f;--colors-displayFgHighestPlaceholder:#767676;--colors-displayFgHighestStrong:#ffffff;--colors-displayFgHighestSelected:#76b900;--colors-displayFgHighestSelectedHover:#8fcb2b;--colors-displayFgHigherPrimary:#f2f2f2;--colors-displayFgHigherSecondary:#c6c6c6;--colors-displayFgHigherDisabled:#8f8f8f;--colors-displayFgHigherPlaceholder:#767676;--colors-displayFgHigherStrong:#ffffff;--colors-displayFgHigherSelected:#76b900;--colors-displayFgHigherSelectedHover:#8fcb2b;--colors-displayFgHighPrimary:#f2f2f2;--colors-displayFgHighSecondary:#c6c6c6;--colors-displayFgHighDisabled:#8f8f8f;--colors-displayFgHighPlaceholder:#767676;--colors-displayFgHighStrong:#ffffff;--colors-displayFgHighSelected:#76b900;--colors-displayFgHighSelectedHover:#8fcb2b;--colors-displayFgMidPrimary:#f2f2f2;--colors-displayFgMidSecondary:#c6c6c6;--colors-displayFgMidDisabled:#8f8f8f;--colors-displayFgMidPlaceholder:#767676;--colors-displayFgMidStrong:#ffffff;--colors-displayFgMidSelected:#76b900;--colors-displayFgMidSelectedHover:#8fcb2b;--colors-displayFgLowPrimary:#f2f2f2;--colors-displayFgLowSecondary:#c6c6c6;--colors-displayFgLowDisabled:#8f8f8f;--colors-displayFgLowPlaceholder:#767676;--colors-displayFgLowStrong:#ffffff;--colors-displayFgLowSelected:#76b900;--colors-displayFgLowSelectedHover:#8fcb2b;--colors-displayFgAccent1:#9accf2;--colors-displayFgAccent1Solid:#001e4b;--colors-displayFgAccent1Bold:#e9f4fb;--colors-displayFgAccent1Disabled:#8f8f8f;--colors-displayFgAccent2:#a7db57;--colors-displayFgAccent2Solid:#23371a;--colors-displayFgAccent2Bold:#ecfad9;--colors-displayFgAccent2Disabled:#8f8f8f;--colors-displayFgAccent3:#e7c32b;--colors-displayFgAccent3Solid:#3a1500;--colors-displayFgAccent3Bold:#f9f4ca;--colors-displayFgAccent3Disabled:#8f8f8f;--colors-displayFgAccent4:#feb3ad;--colors-displayFgAccent4Solid:#480200;--colors-displayFgAccent4Bold:#ffefed;--colors-displayFgAccent4Disabled:#8f8f8f;--colors-displayFgAccent5:#debae8;--colors-displayFgAccent5Solid:#3b0047;--colors-displayFgAccent5Bold:#fbeefe;--colors-displayFgAccent5Disabled:#8f8f8f;--colors-displayFgAccent6:#7cd7c7;--colors-displayFgAccent6Solid:#002522;--colors-displayFgAccent6Bold:#e1f7f3;--colors-displayFgAccent6Disabled:#8f8f8f;--colors-displayFgAccent7:#e2e2e2;--colors-displayFgAccent7Solid:#202020;--colors-displayFgAccent7Bold:#f2f2f2;--colors-displayFgAccent7Disabled:#8f8f8f;--colors-displayFgInverse:#ffffff;--colors-displayFgSubtle:#202020;--colors-displayBorderHighest:#343434;--colors-displayBorderHighestHover:#767676;--colors-displayBorderHighestActive:#c6c6c6;--colors-displayBorderHighestDisabled:#292929;--colors-displayBorderHighestBase:#000000;--colors-displayBorderHighestSelected:#76b900;--colors-displayBorderHighestSelectedHover:#8fcb2b;--colors-displayBorderHigher:#5e5e5e;--colors-displayBorderHigherHover:#8f8f8f;--colors-displayBorderHigherActive:#d8d8d8;--colors-displayBorderHigherDisabled:#292929;--colors-displayBorderHigherBase:#343434;--colors-displayBorderHigherSelected:#76b900;--colors-displayBorderHigherSelectedHover:#8fcb2b;--colors-displayBorderHigh:#5e5e5e;--colors-displayBorderHighHover:#8f8f8f;--colors-displayBorderHighActive:#d8d8d8;--colors-displayBorderHighDisabled:#292929;--colors-displayBorderHighBase:#292929;--colors-displayBorderHighSelected:#76b900;--colors-displayBorderHighSelectedHover:#8fcb2b;--colors-displayBorderMid:#343434;--colors-displayBorderMidHover:#767676;--colors-displayBorderMidActive:#c6c6c6;--colors-displayBorderMidDisabled:#292929;--colors-displayBorderMidBase:#202020;--colors-displayBorderMidSelected:#76b900;--colors-displayBorderMidSelectedHover:#8fcb2b;--colors-displayBorderLow:#343434;--colors-displayBorderLowHover:#767676;--colors-displayBorderLowActive:#c6c6c6;--colors-displayBorderLowDisabled:#292929;--colors-displayBorderLowBase:#121212;--colors-displayBorderLowSelected:#76b900;--colors-displayBorderLowSelectedHover:#8fcb2b;--colors-displayBorderAccent1:#9accf2;--colors-displayBorderAccent1Solid:#9accf2;--colors-displayBorderAccent1SolidHover:#badef8;--colors-displayBorderAccent1SolidActive:#2a63ba;--colors-displayBorderAccent1Bold:#9accf2;--colors-displayBorderAccent1Disabled:#8f8f8f;--colors-displayBorderAccent2:#a7db57;--colors-displayBorderAccent2Solid:#a7db57;--colors-displayBorderAccent2SolidHover:#bfe882;--colors-displayBorderAccent2SolidActive:#57920d;--colors-displayBorderAccent2Bold:#a7db57;--colors-displayBorderAccent2Disabled:#8f8f8f;--colors-displayBorderAccent3:#e7c32b;--colors-displayBorderAccent3Solid:#e7c32b;--colors-displayBorderAccent3SolidHover:#f3d85d;--colors-displayBorderAccent3SolidActive:#924c00;--colors-displayBorderAccent3Bold:#e7c32b;--colors-displayBorderAccent3Disabled:#8f8f8f;--colors-displayBorderAccent4:#feb3ad;--colors-displayBorderAccent4Solid:#feb3ad;--colors-displayBorderAccent4SolidHover:#ffceca;--colors-displayBorderAccent4SolidActive:#b6251b;--colors-displayBorderAccent4Bold:#feb3ad;--colors-displayBorderAccent4Disabled:#8f8f8f;--colors-displayBorderAccent5:#debae8;--colors-displayBorderAccent5Solid:#debae8;--colors-displayBorderAccent5SolidHover:#ead1f3;--colors-displayBorderAccent5SolidActive:#8c3ca0;--colors-displayBorderAccent5Bold:#debae8;--colors-displayBorderAccent5Disabled:#8f8f8f;--colors-displayBorderAccent6:#7cd7c7;--colors-displayBorderAccent6Solid:#7cd7c7;--colors-displayBorderAccent6SolidHover:#a0e6d9;--colors-displayBorderAccent6SolidActive:#006b5e;--colors-displayBorderAccent6Bold:#7cd7c7;--colors-displayBorderAccent6Disabled:#8f8f8f;--colors-displayBorderAccent7:#d8d8d8;--colors-displayBorderAccent7Solid:#c6c6c6;--colors-displayBorderAccent7SolidHover:#d8d8d8;--colors-displayBorderAccent7SolidActive:#5e5e5e;--colors-displayBorderAccent7Bold:#d8d8d8;--colors-displayBorderAccent7Disabled:#8f8f8f;--colors-displayBorderStatic:#5e5e5e;--colors-menuBgDefault:#292929;--colors-menuBgHover:#343434;--colors-menuBgActive:#5e5e5e;--colors-menuBgComplete:#8fcb2b;--colors-menuBgSelected:#202020;--colors-menuBgDisabled:#292929;--colors-menuBgPlaceholder:#292929;--colors-menuBgDanger:#292929;--colors-menuBgDangerHover:#2d0100;--colors-menuBgBase:#000000;--colors-menuBgBaseSelected:#000000;--colors-menuFgDefault:#c6c6c6;--colors-menuFgHover:#f2f2f2;--colors-menuFgActive:#f2f2f2;--colors-menuFgComplete:#f2f2f2;--colors-menuFgSelected:#f2f2f2;--colors-menuFgDisabled:#8f8f8f;--colors-menuFgPlaceholder:#767676;--colors-menuFgDanger:#dc3528;--colors-menuFgDangerHover:#dc3528;--colors-menuFgPrimary:#f2f2f2;--colors-menuFgHeader:#c6c6c6;--colors-menuFgFooter:#c6c6c6;--colors-menuBorderDefault:#343434;--colors-menuBorderHover:#767676;--colors-menuBorderActive:#343434;--colors-menuBorderComplete:#76b900;--colors-menuBorderSelected:#76b900;--colors-menuBorderDisabled:#343434;--colors-menuBorderPlaceholder:#343434;--colors-inputBgDefault:#202020;--colors-inputBgHover:#202020;--colors-inputBgActive:#343434;--colors-inputBgComplete:#000000;--colors-inputBgDisabled:#343434;--colors-inputBgError:#343434;--colors-inputBgSuccess:#000000;--colors-inputFgDefault:#767676;--colors-inputFgHover:#767676;--colors-inputFgActive:#f2f2f2;--colors-inputFgComplete:#f2f2f2;--colors-inputFgDisabled:#8f8f8f;--colors-inputFgError:#f2f2f2;--colors-inputFgErrorMessage:#dc3528;--colors-inputFgSuccess:#f2f2f2;--colors-inputFgSuccessMessage:#76b900;--colors-inputFgLabel:#c6c6c6;--colors-inputFgRequired:#dc3528;--colors-inputFgIcon:#c6c6c6;--colors-inputFgIconHover:#f2f2f2;--colors-inputFgIconActive:#f2f2f2;--colors-inputFgIconComplete:#c6c6c6;--colors-inputFgIconDisabled:#8f8f8f;--colors-inputFgIconError:#dc3528;--colors-inputFgIconSuccess:#76b900;--colors-inputBorderDefault:#8f8f8f;--colors-inputBorderHover:#d8d8d8;--colors-inputBorderActive:#76b900;--colors-inputBorderComplete:#8f8f8f;--colors-inputBorderDisabled:#767676;--colors-inputBorderError:#dc3528;--colors-inputBorderSuccess:#76b900;--colors-actionBgDefault:#292929;--colors-actionBgHover:#121212;--colors-actionBgActive:#000000;--colors-actionBgDisabled:#202020;--colors-actionBgSelected:#23371a;--colors-actionBgSelectedHover:#2f521a;--colors-actionBgSelectedActive:#407015;--colors-actionBgSelectedDisabled:#202020;--colors-actionBgError:#2d0100;--colors-actionBgErrorHover:#480200;--colors-actionBgErrorActive:#6c0d08;--colors-actionBgErrorDisabled:#202020;--colors-actionBgPrimary:#76b900;--colors-actionBgPrimaryHover:#8fcb2b;--colors-actionBgPrimaryActive:#a7db57;--colors-actionBgPrimaryDisabled:#343434;--colors-actionBgSecondary:rgba(255,255,255,0.00);--colors-actionBgSecondaryHover:rgba(255,255,255,0.12);--colors-actionBgSecondaryActive:rgba(255,255,255,0.25);--colors-actionBgSecondaryDisabled:rgba(255,255,255,0.00);--colors-actionBgTertiary:rgba(255,255,255,0.00);--colors-actionBgTertiaryHover:rgba(255,255,255,0.12);--colors-actionBgTertiaryActive:rgba(255,255,255,0.25);--colors-actionBgTertiaryDisabled:rgba(255,255,255,0.00);--colors-actionBgSecondaryInverse:rgba(0,0,0,0.00);--colors-actionBgSecondaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgSecondaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgSecondaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverse:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgTertiaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgTertiaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgPrimaryDanger:#dc3528;--colors-actionBgPrimaryDangerHover:#b6251b;--colors-actionBgPrimaryDangerActive:#6c0d08;--colors-actionBgPrimaryDangerDisabled:#343434;--colors-actionBgSecondaryDanger:rgba(0,0,0,0.00);--colors-actionBgSecondaryDangerHover:#480200;--colors-actionBgSecondaryDangerActive:#6c0d08;--colors-actionBgSecondaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryDanger:rgba(0,0,0,0.00);--colors-actionBgTertiaryDangerHover:#480200;--colors-actionBgTertiaryDangerActive:#6c0d08;--colors-actionBgTertiaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgBase:#5e5e5e;--colors-actionBgBaseDisabled:#343434;--colors-actionBgOn:#76b900;--colors-actionBgOnDisabled:#407015;--colors-actionFgDefault:#f2f2f2;--colors-actionFgHover:#f2f2f2;--colors-actionFgActive:#f2f2f2;--colors-actionFgDisabled:#8f8f8f;--colors-actionFgStrong:#ffffff;--colors-actionFgSelected:#76b900;--colors-actionFgSelectedHover:#57920d;--colors-actionFgSelectedActive:#a7db57;--colors-actionFgSelectedDisabled:#8f8f8f;--colors-actionFgError:#dc3528;--colors-actionFgErrorHover:#b6251b;--colors-actionFgErrorActive:#ec6358;--colors-actionFgErrorDisabled:#8f8f8f;--colors-actionFgPrimary:#202020;--colors-actionFgPrimaryHover:#202020;--colors-actionFgPrimaryActive:#202020;--colors-actionFgPrimaryDisabled:#c6c6c6;--colors-actionFgSecondary:#f2f2f2;--colors-actionFgSecondaryHover:#f2f2f2;--colors-actionFgSecondaryActive:#f2f2f2;--colors-actionFgSecondaryDisabled:#c6c6c6;--colors-actionFgTertiary:#f2f2f2;--colors-actionFgTertiaryHover:#f2f2f2;--colors-actionFgTertiaryActive:#f2f2f2;--colors-actionFgTertiaryDisabled:#c6c6c6;--colors-actionFgSecondaryInverse:#000000;--colors-actionFgSecondaryInverseHover:#000000;--colors-actionFgSecondaryInverseActive:#000000;--colors-actionFgSecondaryInverseDisabled:#767676;--colors-actionFgTertiaryInverse:#000000;--colors-actionFgTertiaryInverseHover:#000000;--colors-actionFgTertiaryInverseActive:#000000;--colors-actionFgTertiaryInverseDisabled:#767676;--colors-actionFgPrimaryDanger:#ffffff;--colors-actionFgPrimaryDangerHover:#ffffff;--colors-actionFgPrimaryDangerActive:#ffffff;--colors-actionFgPrimaryDangerDisabled:#c6c6c6;--colors-actionFgSecondaryDanger:#dc3528;--colors-actionFgSecondaryDangerHover:#ec6358;--colors-actionFgSecondaryDangerActive:#feb3ad;--colors-actionFgSecondaryDangerDisabled:#c6c6c6;--colors-actionFgTertiaryDanger:#dc3528;--colors-actionFgTertiaryDangerHover:#ec6358;--colors-actionFgTertiaryDangerActive:#feb3ad;--colors-actionFgTertiaryDangerDisabled:#c6c6c6;--colors-actionFgInverse:#ffffff;--colors-actionFgOn:#ffffff;--colors-actionFgOnDisabled:#8f8f8f;--colors-actionBorderDefault:#767676;--colors-actionBorderHover:#8f8f8f;--colors-actionBorderActive:#d8d8d8;--colors-actionBorderDisabled:#5e5e5e;--colors-actionBorderSelected:#76b900;--colors-actionBorderSelectedHover:#57920d;--colors-actionBorderSelectedActive:#a7db57;--colors-actionBorderSelectedDisabled:#343434;--colors-actionBorderError:#dc3528;--colors-actionBorderErrorHover:#b6251b;--colors-actionBorderErrorActive:#ec6358;--colors-actionBorderErrorDisabled:#343434;--colors-actionBorderPrimary:#76b900;--colors-actionBorderPrimaryHover:#a7db57;--colors-actionBorderPrimaryActive:#a7db57;--colors-actionBorderPrimaryDisabled:#343434;--colors-actionBorderSecondary:#e2e2e2;--colors-actionBorderSecondaryHover:#ffffff;--colors-actionBorderSecondaryActive:#ffffff;--colors-actionBorderSecondaryDisabled:#c6c6c6;--colors-actionBorderSecondaryInverse:#000000;--colors-actionBorderSecondaryInverseHover:#000000;--colors-actionBorderSecondaryInverseActive:#000000;--colors-actionBorderSecondaryInverseDisabled:#767676;--colors-actionBorderPrimaryDanger:#dc3528;--colors-actionBorderPrimaryDangerHover:#ec6358;--colors-actionBorderPrimaryDangerActive:#ec6358;--colors-actionBorderPrimaryDangerDisabled:#343434;--colors-actionBorderSecondaryDanger:#dc3528;--colors-actionBorderSecondaryDangerHover:#dc3528;--colors-actionBorderSecondaryDangerActive:#ec6358;--colors-actionBorderSecondaryDangerDisabled:#c6c6c6;--colors-actionBorderInverse:#ffffff;--colors-actionBorderBase:#767676;--colors-feedbackBgInfo:#cde6fa;--colors-feedbackBgInfoSolid:#2378ca;--colors-feedbackBgSuccess:#d6f2ad;--colors-feedbackBgSuccessSolid:#76b900;--colors-feedbackBgCaution:#f6e383;--colors-feedbackBgCautionSolid:#e7c32b;--colors-feedbackBgError:#ffdad6;--colors-feedbackBgErrorSolid:#dc3528;--colors-feedbackBgNeutral:#d8d8d8;--colors-feedbackBgProgress:#292929;--colors-feedbackBgTooltip:#343434;--colors-feedbackBgBase:#343434;--colors-feedbackFgInfo:#2378ca;--colors-feedbackFgInfoHover:#2a63ba;--colors-feedbackFgInfoActive:#013076;--colors-feedbackFgSuccess:#76b900;--colors-feedbackFgSuccessHover:#57920d;--colors-feedbackFgSuccessActive:#407015;--colors-feedbackFgCaution:#a96700;--colors-feedbackFgCautionHover:#924c00;--colors-feedbackFgCautionActive:#5a2400;--colors-feedbackFgError:#dc3528;--colors-feedbackFgErrorHover:#b6251b;--colors-feedbackFgErrorActive:#6c0d08;--colors-feedbackFgNeutral:#5e5e5e;--colors-feedbackFgNeutralHover:#343434;--colors-feedbackFgNeutralActive:#000000;--colors-feedbackFgProgress:#76b900;--colors-feedbackFgProgressActive:#407015;--colors-feedbackFgTooltip:#f2f2f2;--colors-feedbackFgPrimary:#f2f2f2;--colors-feedbackFgSecondary:#c6c6c6;--colors-feedbackFgInverse:#ffffff;--colors-feedbackFgReverse:#202020;--colors-feedbackBorderInfo:#2378ca;--colors-feedbackBorderSucces:#76b900;--colors-feedbackBorderCaution:#a96700;--colors-feedbackBorderError:#dc3528;--colors-feedbackBorderNeutral:#8f8f8f;--colors-feedbackBorderProgress:#5e5e5e;--colors-feedbackBorderTooltip:#5e5e5e;--shadows-lightHigher:0px 8px 10px 0 #00000026;--shadows-lightHigh:0px 4px 5px 0 #0000001f;--shadows-lightMid:0px 2px 3px 0 #00000024;--shadows-lightLow:0px 1px 3px 0 #00000033;--shadows-darkHigher:0px 8px 10px 0 #0000004d;--shadows-darkHigh:0px 4px 5px 0 #0000003d;--shadows-darkMid:0px 2px 4px 0 #00000047;--shadows-darkLow:0px 1px 3px 0 #00000066;--shadows-displayModal:0px 8px 10px 0 #0000004d;--shadows-displayPageHeader:0px 2px 4px 0 #00000047;--shadows-displayPopover:0px 4px 5px 0 #0000003d;--shadows-displayCarouselButton:0px 2px 3px 0 #00000024;--shadows-displaySidePanelRight:-4px 0px 5px 0 #00000040;--shadows-displaySidePanelLeft:4px 0px 5px 0 #00000040;--shadows-feedbackTooltip:0px 4px 5px 0 #0000003d;--shadows-feedbackNotification:0px 4px 5px 0 #0000003d;--shadows-menuDropdown:0px 4px 5px 0 #0000003d;--shadows-menuPopout:0px 2px 4px 0 #00000047;--shadows-inputToggle:0px 1px 3px 0 #00000066}.t-gNphrk{--colors-brand:#76b900;--colors-n000:#ffffff;--colors-n025:#fafafa;--colors-n050:#f2f2f2;--colors-n100:#e2e2e2;--colors-n200:#d8d8d8;--colors-n300:#c6c6c6;--colors-n400:#8f8f8f;--colors-n500:#767676;--colors-n600:#5e5e5e;--colors-n700:#343434;--colors-n800:#292929;--colors-n900:#202020;--colors-n950:#121212;--colors-n1000:#000000;--colors-g050:#ecfad9;--colors-g100:#d6f2ad;--colors-g200:#bfe882;--colors-g300:#a7db57;--colors-g400:#8fcb2b;--colors-g500:#76b900;--colors-g600:#57920d;--colors-g700:#407015;--colors-g800:#2f521a;--colors-g900:#23371a;--colors-g950:#192116;--colors-b050:#e9f4fb;--colors-b100:#cde6fa;--colors-b200:#badef8;--colors-b300:#9accf2;--colors-b400:#3a96d9;--colors-b500:#2378ca;--colors-b600:#2a63ba;--colors-b700:#013076;--colors-b800:#00265e;--colors-b900:#001e4b;--colors-b950:#00112c;--colors-y050:#f9f4ca;--colors-y100:#f6e383;--colors-y200:#f3d85d;--colors-y300:#e7c32b;--colors-y400:#c58420;--colors-y500:#a96700;--colors-y600:#924c00;--colors-y700:#5a2400;--colors-y800:#491c00;--colors-y900:#3a1500;--colors-y950:#220c00;--colors-r050:#ffefed;--colors-r100:#ffdad6;--colors-r200:#ffceca;--colors-r300:#feb3ad;--colors-r400:#ec6358;--colors-r500:#dc3528;--colors-r600:#b6251b;--colors-r700:#6c0d08;--colors-r800:#570704;--colors-r900:#480200;--colors-r950:#2d0100;--colors-t050:#e1f7f3;--colors-t100:#b7eee4;--colors-t200:#a0e6d9;--colors-t300:#7cd7c7;--colors-t400:#1fa18d;--colors-t500:#008471;--colors-t600:#006b5e;--colors-t700:#003b35;--colors-t800:#002f2b;--colors-t900:#002522;--colors-t950:#001514;--colors-p050:#fbeefe;--colors-p100:#f1dcf7;--colors-p200:#ead1f3;--colors-p300:#debae8;--colors-p400:#b579c6;--colors-p500:#a05ab4;--colors-p600:#8c3ca0;--colors-p700:#5a066d;--colors-p800:#490457;--colors-p900:#3b0047;--colors-p950:#25002c;--colors-tk000:rgba(0,0,0,0.00);--colors-tk050:rgba(0,0,0,0.05);--colors-tk100:rgba(0,0,0,0.10);--colors-tk120:rgba(0,0,0,0.12);--colors-tk150:rgba(0,0,0,0.15);--colors-tk200:rgba(0,0,0,0.20);--colors-tk300:rgba(0,0,0,0.30);--colors-tk400:rgba(0,0,0,0.40);--colors-tk500:rgba(0,0,0,0.50);--colors-tk600:rgba(0,0,0,0.60);--colors-tk700:rgba(0,0,0,0.70);--colors-tk800:rgba(0,0,0,0.80);--colors-tk900:rgba(0,0,0,0.90);--colors-tw000:rgba(255,255,255,0.00);--colors-tw050:rgba(255,255,255,0.05);--colors-tw100:rgba(255,255,255,0.10);--colors-tw120:rgba(255,255,255,0.12);--colors-tw250:rgba(255,255,255,0.25);--colors-tw200:rgba(255,255,255,0.20);--colors-tw300:rgba(255,255,255,0.30);--colors-tw400:rgba(255,255,255,0.40);--colors-tw500:rgba(255,255,255,0.50);--colors-tw600:rgba(255,255,255,0.60);--colors-tw700:rgba(255,255,255,0.70);--colors-tw800:rgba(255,255,255,0.80);--colors-tw900:rgba(255,255,255,0.90);--colors-displayBgHighest:#ffffff;--colors-displayBgHighestHover:#f2f2f2;--colors-displayBgHighestActive:#e2e2e2;--colors-displayBgHigher:#ffffff;--colors-displayBgHigherHover:#f2f2f2;--colors-displayBgHigherActive:#e2e2e2;--colors-displayBgHigh:#ffffff;--colors-displayBgHighHover:#f2f2f2;--colors-displayBgHighActive:#e2e2e2;--colors-displayBgMid:#ffffff;--colors-displayBgMidHover:#f2f2f2;--colors-displayBgMidActive:#e2e2e2;--colors-displayBgLow:#f2f2f2;--colors-displayBgLowHover:#e2e2e2;--colors-displayBgLowActive:#d8d8d8;--colors-displayBgCanvas:#fafafa;--colors-displayBgAccent1Hover:#e9f4fb;--colors-displayBgAccent1Active:#2a63ba;--colors-displayBgAccent1Solid:#e9f4fb;--colors-displayBgAccent1SolidHover:#cde6fa;--colors-displayBgAccent1SolidActive:#2a63ba;--colors-displayBgAccent1Bold:#2a63ba;--colors-displayBgAccent1Disabled:#8f8f8f;--colors-displayBgAccent2Hover:#ecfad9;--colors-displayBgAccent2Active:#57920d;--colors-displayBgAccent2Solid:#ecfad9;--colors-displayBgAccent2SolidHover:#d6f2ad;--colors-displayBgAccent2SolidActive:#57920d;--colors-displayBgAccent2Bold:#57920d;--colors-displayBgAccent2Disabled:#8f8f8f;--colors-displayBgAccent3Hover:#f9f4ca;--colors-displayBgAccent3Active:#924c00;--colors-displayBgAccent3Solid:#f9f4ca;--colors-displayBgAccent3SolidHover:#f6e383;--colors-displayBgAccent3SolidActive:#924c00;--colors-displayBgAccent3Bold:#924c00;--colors-displayBgAccent3Disabled:#8f8f8f;--colors-displayBgAccent4Hover:#ffefed;--colors-displayBgAccent4Active:#b6251b;--colors-displayBgAccent4Solid:#ffefed;--colors-displayBgAccent4SolidHover:#ffdad6;--colors-displayBgAccent4SolidActive:#b6251b;--colors-displayBgAccent4Bold:#b6251b;--colors-displayBgAccent4Disabled:#8f8f8f;--colors-displayBgAccent5Hover:#fbeefe;--colors-displayBgAccent5Active:#8c3ca0;--colors-displayBgAccent5Solid:#fbeefe;--colors-displayBgAccent5SolidHover:#f1dcf7;--colors-displayBgAccent5SolidActive:#8c3ca0;--colors-displayBgAccent5Bold:#8c3ca0;--colors-displayBgAccent5Disabled:#8f8f8f;--colors-displayBgAccent6Hover:#e1f7f3;--colors-displayBgAccent6Active:#006b5e;--colors-displayBgAccent6Solid:#e1f7f3;--colors-displayBgAccent6SolidHover:#b7eee4;--colors-displayBgAccent6SolidActive:#006b5e;--colors-displayBgAccent6Bold:#006b5e;--colors-displayBgAccent6Disabled:#8f8f8f;--colors-displayBgAccent7Hover:#e2e2e2;--colors-displayBgAccent7Active:#343434;--colors-displayBgAccent7Solid:#d8d8d8;--colors-displayBgAccent7SolidHover:#c6c6c6;--colors-displayBgAccent7SolidActive:#343434;--colors-displayBgAccent7Bold:#343434;--colors-displayBgAccent7Disabled:#8f8f8f;--colors-displayFgHighestPrimary:#202020;--colors-displayFgHighestSecondary:#5e5e5e;--colors-displayFgHighestDisabled:#767676;--colors-displayFgHighestPlaceholder:#8f8f8f;--colors-displayFgHighestStrong:#000000;--colors-displayFgHighestSelected:#76b900;--colors-displayFgHighestSelectedHover:#57920d;--colors-displayFgHigherPrimary:#202020;--colors-displayFgHigherSecondary:#5e5e5e;--colors-displayFgHigherDisabled:#767676;--colors-displayFgHigherPlaceholder:#8f8f8f;--colors-displayFgHigherStrong:#000000;--colors-displayFgHigherSelected:#76b900;--colors-displayFgHigherSelectedHover:#57920d;--colors-displayFgHighPrimary:#202020;--colors-displayFgHighSecondary:#5e5e5e;--colors-displayFgHighDisabled:#767676;--colors-displayFgHighPlaceholder:#8f8f8f;--colors-displayFgHighStrong:#000000;--colors-displayFgHighSelected:#76b900;--colors-displayFgHighSelectedHover:#57920d;--colors-displayFgMidPrimary:#202020;--colors-displayFgMidSecondary:#5e5e5e;--colors-displayFgMidDisabled:#767676;--colors-displayFgMidPlaceholder:#8f8f8f;--colors-displayFgMidStrong:#000000;--colors-displayFgMidSelected:#76b900;--colors-displayFgMidSelectedHover:#57920d;--colors-displayFgLowPrimary:#202020;--colors-displayFgLowSecondary:#5e5e5e;--colors-displayFgLowDisabled:#767676;--colors-displayFgLowPlaceholder:#8f8f8f;--colors-displayFgLowStrong:#000000;--colors-displayFgLowSelected:#76b900;--colors-displayFgLowSelectedHover:#57920d;--colors-displayFgAccent1:#2a63ba;--colors-displayFgAccent1Solid:#2a63ba;--colors-displayFgAccent1Bold:#e9f4fb;--colors-displayFgAccent1Disabled:#8f8f8f;--colors-displayFgAccent2:#57920d;--colors-displayFgAccent2Solid:#57920d;--colors-displayFgAccent2Bold:#ecfad9;--colors-displayFgAccent2Disabled:#8f8f8f;--colors-displayFgAccent3:#924c00;--colors-displayFgAccent3Solid:#924c00;--colors-displayFgAccent3Bold:#f9f4ca;--colors-displayFgAccent3Disabled:#8f8f8f;--colors-displayFgAccent4:#b6251b;--colors-displayFgAccent4Solid:#b6251b;--colors-displayFgAccent4Bold:#ffefed;--colors-displayFgAccent4Disabled:#8f8f8f;--colors-displayFgAccent5:#8c3ca0;--colors-displayFgAccent5Solid:#8c3ca0;--colors-displayFgAccent5Bold:#fbeefe;--colors-displayFgAccent5Disabled:#8f8f8f;--colors-displayFgAccent6:#006b5e;--colors-displayFgAccent6Solid:#006b5e;--colors-displayFgAccent6Bold:#e1f7f3;--colors-displayFgAccent6Disabled:#8f8f8f;--colors-displayFgAccent7:#343434;--colors-displayFgAccent7Solid:#343434;--colors-displayFgAccent7Bold:#f2f2f2;--colors-displayFgAccent7Disabled:#8f8f8f;--colors-displayFgInverse:#ffffff;--colors-displayFgSubtle:#f2f2f2;--colors-displayBorderHighest:#d8d8d8;--colors-displayBorderHighestHover:#8f8f8f;--colors-displayBorderHighestActive:#5e5e5e;--colors-displayBorderHighestDisabled:#e2e2e2;--colors-displayBorderHighestBase:#ffffff;--colors-displayBorderHighestSelected:#76b900;--colors-displayBorderHighestSelectedHover:#57920d;--colors-displayBorderHigher:#d8d8d8;--colors-displayBorderHigherHover:#8f8f8f;--colors-displayBorderHigherActive:#5e5e5e;--colors-displayBorderHigherDisabled:#e2e2e2;--colors-displayBorderHigherBase:#ffffff;--colors-displayBorderHigherSelected:#76b900;--colors-displayBorderHigherSelectedHover:#57920d;--colors-displayBorderHigh:#d8d8d8;--colors-displayBorderHighHover:#8f8f8f;--colors-displayBorderHighActive:#5e5e5e;--colors-displayBorderHighDisabled:#e2e2e2;--colors-displayBorderHighBase:#ffffff;--colors-displayBorderHighSelected:#76b900;--colors-displayBorderHighSelectedHover:#57920d;--colors-displayBorderMid:#d8d8d8;--colors-displayBorderMidHover:#8f8f8f;--colors-displayBorderMidActive:#5e5e5e;--colors-displayBorderMidDisabled:#e2e2e2;--colors-displayBorderMidBase:#ffffff;--colors-displayBorderMidSelected:#76b900;--colors-displayBorderMidSelectedHover:#57920d;--colors-displayBorderLow:#d8d8d8;--colors-displayBorderLowHover:#8f8f8f;--colors-displayBorderLowActive:#5e5e5e;--colors-displayBorderLowDisabled:#e2e2e2;--colors-displayBorderLowBase:#f2f2f2;--colors-displayBorderLowSelected:#76b900;--colors-displayBorderLowSelectedHover:#57920d;--colors-displayBorderAccent1:#2a63ba;--colors-displayBorderAccent1Solid:#e9f4fb;--colors-displayBorderAccent1SolidHover:#cde6fa;--colors-displayBorderAccent1SolidActive:#2a63ba;--colors-displayBorderAccent1Bold:#2a63ba;--colors-displayBorderAccent1Disabled:#8f8f8f;--colors-displayBorderAccent2:#57920d;--colors-displayBorderAccent2Solid:#ecfad9;--colors-displayBorderAccent2SolidHover:#d6f2ad;--colors-displayBorderAccent2SolidActive:#57920d;--colors-displayBorderAccent2Bold:#57920d;--colors-displayBorderAccent2Disabled:#8f8f8f;--colors-displayBorderAccent3:#924c00;--colors-displayBorderAccent3Solid:#f9f4ca;--colors-displayBorderAccent3SolidHover:#f6e383;--colors-displayBorderAccent3SolidActive:#924c00;--colors-displayBorderAccent3Bold:#924c00;--colors-displayBorderAccent3Disabled:#8f8f8f;--colors-displayBorderAccent4:#b6251b;--colors-displayBorderAccent4Solid:#ffefed;--colors-displayBorderAccent4SolidHover:#ffdad6;--colors-displayBorderAccent4SolidActive:#b6251b;--colors-displayBorderAccent4Bold:#b6251b;--colors-displayBorderAccent4Disabled:#8f8f8f;--colors-displayBorderAccent5:#8c3ca0;--colors-displayBorderAccent5Solid:#fbeefe;--colors-displayBorderAccent5SolidHover:#f1dcf7;--colors-displayBorderAccent5SolidActive:#8c3ca0;--colors-displayBorderAccent5Bold:#8c3ca0;--colors-displayBorderAccent5Disabled:#8f8f8f;--colors-displayBorderAccent6:#006b5e;--colors-displayBorderAccent6Solid:#e1f7f3;--colors-displayBorderAccent6SolidHover:#b7eee4;--colors-displayBorderAccent6SolidActive:#006b5e;--colors-displayBorderAccent6Bold:#006b5e;--colors-displayBorderAccent6Disabled:#8f8f8f;--colors-displayBorderAccent7:#343434;--colors-displayBorderAccent7Solid:#d8d8d8;--colors-displayBorderAccent7SolidHover:#c6c6c6;--colors-displayBorderAccent7SolidActive:#343434;--colors-displayBorderAccent7Bold:#343434;--colors-displayBorderAccent7Disabled:#8f8f8f;--colors-displayBorderStatic:#c6c6c6;--colors-menuBgDefault:#ffffff;--colors-menuBgHover:#f2f2f2;--colors-menuBgActive:#e2e2e2;--colors-menuBgComplete:#8fcb2b;--colors-menuBgSelected:#f2f2f2;--colors-menuBgDisabled:#ffffff;--colors-menuBgPlaceholder:#ffffff;--colors-menuBgDanger:#ffffff;--colors-menuBgDangerHover:#ffefed;--colors-menuBgBase:#ffffff;--colors-menuBgBaseSelected:#ffffff;--colors-menuFgDefault:#5e5e5e;--colors-menuFgHover:#202020;--colors-menuFgActive:#202020;--colors-menuFgComplete:#202020;--colors-menuFgSelected:#202020;--colors-menuFgDisabled:#767676;--colors-menuFgPlaceholder:#8f8f8f;--colors-menuFgDanger:#dc3528;--colors-menuFgDangerHover:#dc3528;--colors-menuFgPrimary:#202020;--colors-menuFgHeader:#5e5e5e;--colors-menuFgFooter:#5e5e5e;--colors-menuBorderDefault:#d8d8d8;--colors-menuBorderHover:#8f8f8f;--colors-menuBorderActive:#d8d8d8;--colors-menuBorderComplete:#76b900;--colors-menuBorderSelected:#76b900;--colors-menuBorderDisabled:#d8d8d8;--colors-menuBorderPlaceholder:#d8d8d8;--colors-inputBgDefault:#ffffff;--colors-inputBgHover:#ffffff;--colors-inputBgActive:#f2f2f2;--colors-inputBgComplete:#ffffff;--colors-inputBgDisabled:#f2f2f2;--colors-inputBgError:#f2f2f2;--colors-inputBgSuccess:#ffffff;--colors-inputFgDefault:#8f8f8f;--colors-inputFgHover:#8f8f8f;--colors-inputFgActive:#202020;--colors-inputFgComplete:#202020;--colors-inputFgDisabled:#767676;--colors-inputFgError:#202020;--colors-inputFgErrorMessage:#dc3528;--colors-inputFgSuccess:#202020;--colors-inputFgSuccessMessage:#76b900;--colors-inputFgLabel:#5e5e5e;--colors-inputFgRequired:#dc3528;--colors-inputFgIcon:#5e5e5e;--colors-inputFgIconHover:#202020;--colors-inputFgIconActive:#202020;--colors-inputFgIconComplete:#5e5e5e;--colors-inputFgIconDisabled:#767676;--colors-inputFgIconError:#dc3528;--colors-inputFgIconSuccess:#76b900;--colors-inputBorderDefault:#c6c6c6;--colors-inputBorderHover:#767676;--colors-inputBorderActive:#76b900;--colors-inputBorderComplete:#c6c6c6;--colors-inputBorderDisabled:#d8d8d8;--colors-inputBorderError:#dc3528;--colors-inputBorderSuccess:#76b900;--colors-actionBgDefault:#ffffff;--colors-actionBgHover:#e2e2e2;--colors-actionBgActive:#c6c6c6;--colors-actionBgDisabled:#f2f2f2;--colors-actionBgSelected:#ecfad9;--colors-actionBgSelectedHover:#d6f2ad;--colors-actionBgSelectedActive:#a7db57;--colors-actionBgSelectedDisabled:#f2f2f2;--colors-actionBgError:#ffefed;--colors-actionBgErrorHover:#ffdad6;--colors-actionBgErrorActive:#feb3ad;--colors-actionBgErrorDisabled:#f2f2f2;--colors-actionBgPrimary:#76b900;--colors-actionBgPrimaryHover:#8fcb2b;--colors-actionBgPrimaryActive:#a7db57;--colors-actionBgPrimaryDisabled:#d8d8d8;--colors-actionBgSecondary:rgba(0,0,0,0.00);--colors-actionBgSecondaryHover:rgba(0,0,0,0.05);--colors-actionBgSecondaryActive:rgba(0,0,0,0.15);--colors-actionBgSecondaryDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiary:rgba(0,0,0,0.00);--colors-actionBgTertiaryHover:rgba(0,0,0,0.05);--colors-actionBgTertiaryActive:rgba(0,0,0,0.15);--colors-actionBgTertiaryDisabled:rgba(0,0,0,0.00);--colors-actionBgSecondaryInverse:rgba(0,0,0,0.00);--colors-actionBgSecondaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgSecondaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgSecondaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverse:rgba(0,0,0,0.00);--colors-actionBgTertiaryInverseHover:rgba(0,0,0,0.10);--colors-actionBgTertiaryInverseActive:rgba(0,0,0,0.20);--colors-actionBgTertiaryInverseDisabled:rgba(0,0,0,0.00);--colors-actionBgPrimaryDanger:#dc3528;--colors-actionBgPrimaryDangerHover:#b6251b;--colors-actionBgPrimaryDangerActive:#570704;--colors-actionBgPrimaryDangerDisabled:#d8d8d8;--colors-actionBgSecondaryDanger:rgba(0,0,0,0.00);--colors-actionBgSecondaryDangerHover:#ffdad6;--colors-actionBgSecondaryDangerActive:#feb3ad;--colors-actionBgSecondaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgTertiaryDanger:rgba(0,0,0,0.00);--colors-actionBgTertiaryDangerHover:#ffdad6;--colors-actionBgTertiaryDangerActive:#feb3ad;--colors-actionBgTertiaryDangerDisabled:rgba(0,0,0,0.00);--colors-actionBgBase:#c6c6c6;--colors-actionBgBaseDisabled:#e2e2e2;--colors-actionBgOn:#76b900;--colors-actionBgOnDisabled:#a7db57;--colors-actionFgDefault:#202020;--colors-actionFgHover:#202020;--colors-actionFgActive:#202020;--colors-actionFgDisabled:#767676;--colors-actionFgStrong:#000000;--colors-actionFgSelected:#76b900;--colors-actionFgSelectedHover:#57920d;--colors-actionFgSelectedActive:#407015;--colors-actionFgSelectedDisabled:#767676;--colors-actionFgError:#dc3528;--colors-actionFgErrorHover:#b6251b;--colors-actionFgErrorActive:#570704;--colors-actionFgErrorDisabled:#767676;--colors-actionFgPrimary:#202020;--colors-actionFgPrimaryHover:#202020;--colors-actionFgPrimaryActive:#202020;--colors-actionFgPrimaryDisabled:#767676;--colors-actionFgSecondary:#202020;--colors-actionFgSecondaryHover:#202020;--colors-actionFgSecondaryActive:#202020;--colors-actionFgSecondaryDisabled:#8f8f8f;--colors-actionFgTertiary:#202020;--colors-actionFgTertiaryHover:#202020;--colors-actionFgTertiaryActive:#202020;--colors-actionFgTertiaryDisabled:#8f8f8f;--colors-actionFgSecondaryInverse:#ffffff;--colors-actionFgSecondaryInverseHover:#ffffff;--colors-actionFgSecondaryInverseActive:#ffffff;--colors-actionFgSecondaryInverseDisabled:#d8d8d8;--colors-actionFgTertiaryInverse:#ffffff;--colors-actionFgTertiaryInverseHover:#ffffff;--colors-actionFgTertiaryInverseActive:#ffffff;--colors-actionFgTertiaryInverseDisabled:#d8d8d8;--colors-actionFgPrimaryDanger:#ffffff;--colors-actionFgPrimaryDangerHover:#ffffff;--colors-actionFgPrimaryDangerActive:#ffffff;--colors-actionFgPrimaryDangerDisabled:#767676;--colors-actionFgSecondaryDanger:#dc3528;--colors-actionFgSecondaryDangerHover:#b6251b;--colors-actionFgSecondaryDangerActive:#6c0d08;--colors-actionFgSecondaryDangerDisabled:#8f8f8f;--colors-actionFgTertiaryDanger:#dc3528;--colors-actionFgTertiaryDangerHover:#b6251b;--colors-actionFgTertiaryDangerActive:#6c0d08;--colors-actionFgTertiaryDangerDisabled:#8f8f8f;--colors-actionFgInverse:#ffffff;--colors-actionFgOn:#ffffff;--colors-actionFgOnDisabled:#f2f2f2;--colors-actionBorderDefault:#8f8f8f;--colors-actionBorderHover:#767676;--colors-actionBorderActive:#343434;--colors-actionBorderDisabled:#c6c6c6;--colors-actionBorderSelected:#76b900;--colors-actionBorderSelectedHover:#57920d;--colors-actionBorderSelectedActive:#407015;--colors-actionBorderSelectedDisabled:#d8d8d8;--colors-actionBorderError:#dc3528;--colors-actionBorderErrorHover:#b6251b;--colors-actionBorderErrorActive:#570704;--colors-actionBorderErrorDisabled:#d8d8d8;--colors-actionBorderPrimary:#76b900;--colors-actionBorderPrimaryHover:#a7db57;--colors-actionBorderPrimaryActive:#a7db57;--colors-actionBorderPrimaryDisabled:#d8d8d8;--colors-actionBorderSecondary:#5e5e5e;--colors-actionBorderSecondaryHover:#202020;--colors-actionBorderSecondaryActive:#202020;--colors-actionBorderSecondaryDisabled:#8f8f8f;--colors-actionBorderSecondaryInverse:#ffffff;--colors-actionBorderSecondaryInverseHover:#ffffff;--colors-actionBorderSecondaryInverseActive:#ffffff;--colors-actionBorderSecondaryInverseDisabled:#d8d8d8;--colors-actionBorderPrimaryDanger:#dc3528;--colors-actionBorderPrimaryDangerHover:#570704;--colors-actionBorderPrimaryDangerActive:#570704;--colors-actionBorderPrimaryDangerDisabled:#d8d8d8;--colors-actionBorderSecondaryDanger:#dc3528;--colors-actionBorderSecondaryDangerHover:#dc3528;--colors-actionBorderSecondaryDangerActive:#6c0d08;--colors-actionBorderSecondaryDangerDisabled:#8f8f8f;--colors-actionBorderInverse:#ffffff;--colors-actionBorderBase:#8f8f8f;--colors-feedbackBgInfo:#e9f4fb;--colors-feedbackBgInfoSolid:#2378ca;--colors-feedbackBgSuccess:#ecfad9;--colors-feedbackBgSuccessSolid:#76b900;--colors-feedbackBgCaution:#f9f4ca;--colors-feedbackBgCautionSolid:#e7c32b;--colors-feedbackBgError:#ffefed;--colors-feedbackBgErrorSolid:#dc3528;--colors-feedbackBgNeutral:#e2e2e2;--colors-feedbackBgProgress:#ffffff;--colors-feedbackBgTooltip:#343434;--colors-feedbackBgBase:#ffffff;--colors-feedbackFgInfo:#2378ca;--colors-feedbackFgInfoHover:#2a63ba;--colors-feedbackFgInfoActive:#013076;--colors-feedbackFgSuccess:#76b900;--colors-feedbackFgSuccessHover:#57920d;--colors-feedbackFgSuccessActive:#407015;--colors-feedbackFgCaution:#a96700;--colors-feedbackFgCautionHover:#924c00;--colors-feedbackFgCautionActive:#5a2400;--colors-feedbackFgError:#dc3528;--colors-feedbackFgErrorHover:#b6251b;--colors-feedbackFgErrorActive:#6c0d08;--colors-feedbackFgNeutral:#5e5e5e;--colors-feedbackFgNeutralHover:#343434;--colors-feedbackFgNeutralActive:#000000;--colors-feedbackFgProgress:#76b900;--colors-feedbackFgProgressActive:#407015;--colors-feedbackFgTooltip:#f2f2f2;--colors-feedbackFgPrimary:#202020;--colors-feedbackFgSecondary:#5e5e5e;--colors-feedbackFgInverse:#ffffff;--colors-feedbackFgReverse:#202020;--colors-feedbackBorderInfo:#2378ca;--colors-feedbackBorderSuccess:#76b900;--colors-feedbackBorderCaution:#a96700;--colors-feedbackBorderError:#dc3528;--colors-feedbackBorderNeutral:#8f8f8f;--colors-feedbackBorderProgress:#d8d8d8;--colors-feedbackBorderTooltip:#d8d8d8;--shadows-lightHigher:0px 8px 10px 0 #00000026;--shadows-lightHigh:0px 4px 5px 0 #0000001f;--shadows-lightMid:0px 2px 3px 0 #00000024;--shadows-lightLow:0px 1px 3px 0 #00000033;--shadows-darkHigher:0px 8px 10px 0 #0000004d;--shadows-darkHigh:0px 4px 5px 0 #0000003d;--shadows-darkMid:0px 2px 4px 0 #00000047;--shadows-darkLow:0px 1px 3px 0 #00000066;--shadows-displayModal:0px 8px 10px 0 #00000026;--shadows-displayPageHeader:0px 2px 3px 0 #00000024;--shadows-displayPopover:0px 4px 5px 0 #0000001f;--shadows-displayCarouselButton:0px 2px 3px 0 #00000024;--shadows-displaySidePanelRight:-4px 0px 5px 0 #0000001f;--shadows-displaySidePanelLeft:4px 0px 5px 0 #0000001f;--shadows-feedbackTooltip:0px 4px 5px 0 #0000001f;--shadows-feedbackNotification:0px 4px 5px 0 #0000001f;--shadows-menuDropdown:0px 4px 5px 0 #0000001f;--shadows-menuPopout:0px 2px 3px 0 #00000024;--shadows-inputToggle:0px 1px 3px 0 #00000033}}--sxs{--sxs:1 hYqCjV}@media{[data-radix-popper-content-wrapper]{z-index:1060 !important}}--sxs{--sxs:2 c-dwAUIm c-gsmDXe c-jmVDLn c-gVWXCF c-dFMixa c-PJLV c-gHdmVn c-coDgbF c-hakyQ c-dhzjXW c-dMjTYf c-fZrNKq c-TKGAm c-ilGjSy c-iRRrYf c-dXKYtZ c-dyDYjC c-jUxYmc c-Dadvv c-cyMReR c-cULKyM c-fpDrvI c-tvOOo c-cmpvrW c-cxpKzu c-UazGY c-fyVBxt c-jOMmSw c-fEzUaR c-enuzaa c-dSpuOG c-kAArxJ c-duhgWy c-dWeQby c-dCzjnn c-fHYEhm c-ijtrue c-kMaXzg c-eFiwTo c-cQcYE c-kzncD c-gslCyW c-ElNdM c-MCGcG c-kqPaQQ c-jTLszv c-gWUlrN c-fVhFNE c-WUEIO c-cKCGFj c-hPmoew c-dGiVjH c-buhBet c-gGIOsy c-cqgeFx c-iZqJcm c-iaIqNA c-iGJBGr c-epPExb c-cwlcYD c-hmOgCM c-kYsbuo c-gRrrJB c-jEISYs c-cQehLr c-dsPRoj c-jgxVxd}@media{.c-dwAUIm{display:block;background-color:var(--colors-displayBgCanvas);color:var(--colors-displayFgLowPrimary);font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);line-height:var(--lineHeights-body)}.c-gsmDXe{cursor:pointer}.c-jmVDLn{position:fixed;top:50%;left:50%;transform:translate(-50%, -50%);z-index:1050}.c-gVWXCF{box-sizing:border-box;font-family:var(--fonts-nvidiaSans);line-height:var(--lineHeights-body);background-color:var(--colors-displayBgMid);border:1px solid transparent;color:var(--colors-displayFgLowPrimary);display:grid;grid-template-areas: 'modal-title' 'modal-content' 'modal-footer' ;row-gap:var(--space-md);box-shadow:var(--shadows-displayModal);border-radius:var(--borderRadius-md)}.c-dFMixa{background-color:hsla(0, 0%, 0%, 0.40);position:fixed;inset:0;animation:overlayShow 150ms cubic-bezier(0.16, 1, 0.3, 1);z-index:1040}.c-gHdmVn{background-color:var(--colors-menuBgBase);display:inline-flex;flex-direction:column;place-content:center;border:1px solid var(--colors-menuBorderDefault);box-shadow:var(--shadows-menuDropdown);border-radius:var(--borderRadius-md)}.c-coDgbF{overflow-y:auto}.c-hakyQ{display:block}.c-dhzjXW{display:flex}.c-dMjTYf{box-sizing:border-box;font-family:var(--fonts-nvidiaSans);line-height:var(--lineHeights-body)}.c-fZrNKq{display:flex;width:100%}.c-TKGAm{cursor:pointer;border:1px solid var(--colors-g500);display:flex;justify-content:center;align-items:center;gap:var(--space-xs);font-size:var(--fontSizes-ms);font-weight:var(--fontWeights-medium);font-family:var(--fonts-nvidiaSans);line-height:var(--lineHeights-text);padding:var(--space-sm) var(--space-ms)}.c-TKGAm:disabled:hover{cursor:not-allowed}.c-ilGjSy{position:relative;background-color:transparent;color:var(--colors-menuFgPrimary);display:flex;flex-direction:row;flex-wrap:nowrap;overflow:hidden;width:100%}.c-ilGjSy:before{background-color:var(--colors-menuBorderDefault);height:2px;content:"";display:block;inset:auto 0 0 0;position:absolute}.c-iRRrYf{display:flex;flex-grow:0;flex-shrink:1;overflow-x:auto;position:relative;scroll-snap-type:x mandatory}.c-iRRrYf::-webkit-scrollbar{display:none}.c-dXKYtZ{position:relative;overflow-x:auto}.c-dXKYtZ::-webkit-scrollbar{display:none}.c-dyDYjC{---gradientColor:var(--colors-menuBgDefault);height:100%;position:absolute;width:100px;z-index:10;opacity:1;transition:opacity 0.25s ease-in-out}.c-jUxYmc{background:none;border:0 none;cursor:pointer;font-family:var(--fonts-nvidiaSans);font-weight:var(--fontWeights-regular);line-height:var(--lineHeights-text);font-size:var(--fontSizes-ms);letter-spacing:var(--letterSpacings-none);text-case:none;text-decoration:none;paragraph-spacing:0;position:relative;flex:0 0 auto;color:var(--colors-menuFgPrimary);padding:var(--space-ms)}.c-jUxYmc:before{background-color:transparent;height:2px;content:"";display:block;inset:auto 0 0 0;position:absolute}.c-jUxYmc:hover:not([data-disabled]):not([data-state="active"]):not([ng-reflect-is-active="true"]):not([ng-reflect-is-disabled="true"]){background-color:transparent;border-color:var(--colors-menuBorderHover);color:var(--colors-menuFgHover)}.c-jUxYmc:hover:not([data-disabled]):not([data-state="active"]):not([ng-reflect-is-active="true"]):not([ng-reflect-is-disabled="true"]):before{background-color:var(--colors-menuBorderHover);height:2px}.c-jUxYmc:active:not([data-disabled]){background-color:var(--colors-menuBgActive);border-color:var(--colors-menuBorderActive);color:var(--colors-menuFgActive)}.c-jUxYmc:active:not([data-disabled]):before{background-color:var(--colors-menuBorderSelected);height:4px}.c-Dadvv{fill:var(--colors-feedbackFgProgress);width:0;height:0;animation:k-iOdNKk 1s infinite}@media only screen and (max-width: 1023px){.c-cyMReR .output-slot{min-width:500px}}@media screen and (max-width: 639px){.c-cULKyM{display:none}}.c-fpDrvI{font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-sm);font-weight:var(--fontWeights-medium);line-height:var(--lineHeights-text);display:flex;align-items:baseline;gap:var(--space-xs);color:var(--colors-inputFgLabel)}.c-tvOOo{color:var(--colors-inputFgRequired)}.c-cmpvrW{position:relative}.c-cxpKzu{box-sizing:border-box;border:var(--borderWidths-sm) solid var(--colors-inputBorderDefault);border-radius:var(--borderRadius-sm);width:100%;display:flex;justify-content:space-between;align-items:center;font-family:var(--fonts-nvidiaSans);line-height:var(--lineHeights-body);background-color:var(--colors-inputBgDefault);color:var(--colors-inputFgComplete)}.c-cxpKzu:hover{border-color:var(--colors-inputFgHover)}.c-cxpKzu[data-state="open"]{border-image:linear-gradient(to top, var(--colors-inputBorderActive), var(--colors-inputBorderActive) 2px, transparent 2px, transparent) 2;outline:none;background-color:var(--colors-inputBgActive)}.c-cxpKzu[data-disabled]{background-color:var(--colors-inputBgDisabled);border-color:var(--colors-inputBorderDisabled);color:var(--colors-inputFgDisabled);cursor:not-allowed}.c-cxpKzu[data-disabled]:hover{border-color:var(--colors-inputBorderDisabled)}.c-cxpKzu[data-placeholder]{color:var(--colors-inputFgDefault)}.c-cxpKzu > div:first-child{min-width:0px}.c-cxpKzu > div:first-child > span{overflow:hidden;white-space:nowrap;text-overflow:ellipsis}.c-UazGY{display:flex;align-items:center}.c-fyVBxt{box-sizing:border-box;border:var(--borderWidths-sm) solid var(--colors-inputBorderDefault);background-color:var(--colors-inputBgDefault);color:var(--colors-inputFgComplete);font-size:var(--fontSizes-ms);line-height:var(--lineHeights-body);border-radius:var(--borderRadius-sm)}.c-fyVBxt:hover{border-color:var(--colors-inputFgHover)}.c-fyVBxt:focus-within{border-image:linear-gradient(to top, var(--colors-inputBorderActive), var(--colors-inputBorderActive) 2px, transparent 2px, transparent) 2;outline:none;background-color:var(--colors-inputBgActive)}.c-jOMmSw{all:unset;border:none;box-sizing:border-box;color:var(--colors-inputFgComplete);display:block;width:100%;font-family:var(--fonts-nvidiaSans);font-size:inherit}.c-jOMmSw::placeholder{color:var(--colors-inputFgDefault)}.c-jOMmSw:focus{outline:none}.c-fEzUaR{border:0 solid #5e5e5e;border-radius:10px}.c-fEzUaR:has(input:checked),.c-fEzUaR:focus-within{border:0 solid #76b900 !important}.c-fEzUaR:hover:not(:has(input:checked)){border-color:#8f8f8f}.c-fEzUaR{display:block;position:relative}.c-enuzaa{position:absolute;opacity:0;width:100%;height:100%}.c-enuzaa:hover{cursor:pointer}.c-dSpuOG{-webkit-appearance:none;appearance:none;border:var(--borderWidths-md) solid var(--colors-actionBorderDefault);border-radius:50%;display:grid;height:var(--space-md);place-content:center;width:var(--space-md);background-color:var(--colors-actionBgDefault)}.c-dSpuOG:hover:enabled{background-color:var(--colors-actionBgHover);border-color:var(--colors-actionBorderHover)}.c-dSpuOG:active:enabled{background-color:var(--colors-actionBgActive);border-color:var(--colors-actionBorderActive)}.c-dSpuOG::before{border-radius:50%;box-shadow:inset 1em 1em var(--colors-actionFgSelected);content:"";height:var(--space-sm);transform:scale(0);width:var(--space-sm);background-color:CanvasText}.c-dSpuOG:checked{background-color:var(--colors-actionBgSelected);border-color:var(--colors-actionBorderSelected)}.c-dSpuOG:checked::before{transform:scale(1);transition:120ms transform ease-in-out}.c-dSpuOG:checked:hover:enabled{background-color:var(--colors-actionBgSelectedHover);border-color:var(--colors-actionBorderSelectedHover)}.c-dSpuOG:checked:hover:enabled::before{box-shadow:inset 1em 1em var(--colors-actionFgSelectedHover)}.c-dSpuOG:checked:active:enabled{background-color:var(--colors-actionBgSelectedActive);border-color:var(--colors-actionBorderSelectedActive)}.c-dSpuOG:checked:active:enabled::before{box-shadow:inset 1em 1em var(--colors-actionFgSelectedActive)}.c-dSpuOG:disabled{background-color:var(--colors-actionBgDisabled);border-color:var(--colors-actionBorderDisabled)}.c-dSpuOG:disabled:checked{background-color:var(--colors-actionBgSelectedDisabled);border-color:var(--colors-actionBorderDisabled)}.c-dSpuOG:disabled:checked::before{box-shadow:inset 1em 1em var(--colors-actionFgDisabled)}.c-dSpuOG:disabled:hover{cursor:not-allowed}.c-kAArxJ{---fontFamily:var(--fonts-nvidiaSans);---fontSize:var(--fontSizes-ms);---fontWeight:var(--fontWeights-regular);---lineHeight:var(--lineHeights-text);font-family:var(---fontFamily);font-size:var(---fontSize);font-weight:var(---fontWeight);line-height:var(---lineHeight)}.c-duhgWy{box-sizing:border-box;font-family:var(--fonts-nvidiaSans);position:relative;display:flex;align-items:center;-webkit-user-select:none;user-select:none;touch-action:none}.c-duhgWy[data-orientation="horizontal"]{height:var(--space-ml)}.c-duhgWy[data-orientation="vertical"]{width:var(--space-ml);height:100%;min-height:350px;flex-direction:column}.c-dWeQby{background-color:var(--colors-actionBgBase);position:relative;flex-grow:1;border-radius:9999px}.c-dWeQby[data-disabled]{background-color:var(--colors-actionBgBaseDisabled)}.c-dWeQby[data-orientation="horizontal"]{height:var(--space-xs);width:100%}.c-dWeQby[data-orientation="vertical"]{height:100%;width:var(--space-xs)}.c-dCzjnn{position:absolute;background-color:var(--colors-actionBgOn);border-radius:9999px}.c-dCzjnn[data-disabled]{background-color:var(--colors-actionBgOnDisabled)}.c-dCzjnn[data-orientation="horizontal"]{height:100%}.c-dCzjnn[data-orientation="vertical"]{width:100%}.c-fHYEhm{display:block;width:var(--space-ml);height:var(--space-ml);background-color:var(--colors-inputBgDefault);transform:matrix(1, 0, 0, -1, 0, 0);border:3px solid var(--colors-actionBorderBase);border-radius:50%}.c-fHYEhm:focus{outline:none;box-shadow:var(--shadows-darkLow)}.c-fHYEhm:hover{background-color:var(--colors-inputBgActive)}.c-fHYEhm[data-disabled]{background-color:var(--colors-inputBgDisabled);border-color:var(--colors-inputBorderDisabled)}.c-ijtrue{position:relative;display:flex}.c-ijtrue[data-orientation="horizontal"]{margin-top:var(--space-ss);width:unset}.c-ijtrue[data-orientation="vertical"]{position:absolute;transform:translateY(-100%)}.c-kMaXzg{display:inline-flex;align-items:center;position:absolute}.c-kMaXzg[data-orientation="horizontal"]{flex-direction:column;transform:translateX(-50%)}.c-kMaXzg[data-orientation="vertical"]{flex-direction:row;transform:translateY(50%)}.c-kMaXzg[data-orientation="vertical"] .c-SfqdT{margin-right:var(--space-ss)}.c-eFiwTo{color:var(--colors-inputFgLabel);font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-sm);font-weight:var(--fontWeights-regular)}.c-cQcYE{border:var(--borderWidths-sm) solid var(--colors-inputBorderDefault);box-sizing:border-box;position:relative;background-color:var(--colors-inputBgDefault);color:var(--colors-inputFgComplete);border-radius:var(--borderRadius-sm)}.c-cQcYE:hover{border-color:var(--colors-inputFgHover)}.c-cQcYE:focus-within{border-image:linear-gradient(to top, var(--colors-inputBorderActive), var(--colors-inputBorderActive) 2px, transparent 2px, transparent) 2;outline:none;background-color:var(--colors-inputBgActive)}.c-kzncD{all:unset;border:none;box-sizing:border-box;display:block;height:100%;width:100%;font-family:var(--fonts-nvidiaSans);line-height:var(--lineHeights-text);text-align:left}.c-kzncD::placeholder{color:var(--colors-inputFgDefault)}.c-kzncD:focus{outline:none}.c-gslCyW{display:block;background-color:var(--colors-displayBgHigh);border-radius:var(--borderRadius-md)}.c-ElNdM{display:block;border-bottom:1px solid var(--colors-displayBorderHigh)}.c-ElNdM:last-child{border-bottom-right-radius:var(--borderRadius-md);border-bottom-left-radius:var(--borderRadius-md)}.c-MCGcG{background:none;border:0 none;cursor:pointer;font-size:initial;text-align:left;align-items:center;display:flex;gap:var(--space-md);min-length:0;padding:var(--space-ms) var(--space-md);width:100%;color:var(--colors-displayFgHighPrimary)}.c-MCGcG:hover:not(:disabled){background-color:var(--colors-displayBgHighHover)}.c-MCGcG:disabled{cursor:not-allowed;color:var(--colors-displayFgHighDisabled)}.c-kqPaQQ{display:inline-flex;justify-content:center;align-items:center;flex:0 0 auto}.c-jTLszv{flex:1 1 auto;min-width:0;overflow:hidden;text-overflow:ellipsis;white-space:nowrap}.c-gWUlrN{font-weight:var(--fontWeights-medium);font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms)}.c-gWUlrN abbr{text-decoration:none}.c-fVhFNE{display:block;padding:var(--space-ms) var(--space-md)}.c-WUEIO{height:22px;display:flex;column-gap:var(--space-ss);row-gap:var(--space-xs);padding:0 !important}.c-cKCGFj{padding:0 !important;margin:0 !important;color:var(--colors-inputFgComplete) !important}@media screen and (max-width: 851px){.c-hPmoew{display:none}}.c-dGiVjH{background-color:#121212;border:1px solid #5E5E5E;padding:var(--space-ms);font-family:var(--fonts-jetbrainsMono);border-radius:var(--space-sm)}.c-buhBet{-webkit-appearance:none;appearance:none;border:var(--borderWidths-md) solid var(--colors-actionBorderDefault);border-radius:var(--borderRadius-sm);display:grid;height:var(--space-md);place-content:center;width:var(--space-md);background-color:var(--colors-actionBgDefault)}.c-buhBet:hover:enabled{background-color:var(--colors-actionBgHover);border-color:var(--colors-actionBorderHover)}.c-buhBet:active:enabled{background-color:var(--colors-actionBgActive);border-color:var(--colors-actionBorderActive)}.c-buhBet::before{box-shadow:inset 1em 1em var(--colors-actionBorderInverse);-webkit-clip-path:polygon(14.97% 57.84%, 24.71% 47.86%, 37.78% 60.63%, 77.61% 19.83%, 87.36% 29.35%, 37.78% 80.12%);clip-path:polygon(14.97% 57.84%, 24.71% 47.86%, 37.78% 60.63%, 77.61% 19.83%, 87.36% 29.35%, 37.78% 80.12%);content:"";height:var(--space-md);transform:scale(0);width:var(--space-md);background-color:CanvasText}.c-buhBet:checked{background-color:var(--colors-actionFgSelected);border:none}.c-buhBet:checked::before{transform:scale(1)}.c-buhBet:checked:hover:enabled{background-color:var(--colors-actionFgSelectedHover)}.c-buhBet:checked:active:enabled{background-color:var(--colors-actionFgSelectedActive)}.c-buhBet:disabled{background-color:var(--colors-actionBgDisabled);border-color:var(--colors-actionBorderDisabled)}.c-buhBet:disabled:checked{background-color:var(--colors-actionFgSelectedDisabled)}.c-buhBet:disabled:checked::before{box-shadow:inset 1em 1em var(--colors-actionBorderDisabled)}.c-buhBet:disabled:hover{cursor:not-allowed}.c-buhBet:indeterminate{background-color:var(--colors-actionFgSelected)}.c-buhBet:indeterminate:hover{background-color:var(--colors-actionFgSelectedHover)}.c-buhBet:indeterminate:active{background-color:var(--colors-actionFgSelectedActive)}.c-buhBet:indeterminate::before{content:"";display:block;position:relative}.c-buhBet:indeterminate::after{background-color:var(--colors-actionFgSelected);content:"";display:block;border:solid var(--colors-actionFgOn);border-width:2px 0 0 0;width:8px;position:absolute;top:23px;left:20px}.c-gGIOsy{color:var(--colors-feedbackFgPrimary);letter-spacing:var(--letterSpacings-less);margin-top:var(--space-md);text-align:center}.c-cqgeFx{color:var(--colors-feedbackFgSecondary);font-size:var(--fontSizes-ms);font-weight:var(--fontWeights-regular);text-align:center;line-height:var(--lineHeights-body);margin-top:var(--space-xxs)}.c-iZqJcm{margin-top:var(--space-md);gap:var(--space-sm)}.c-iaIqNA > path:first-of-type{fill:#dc3528}.c-iGJBGr{gap:0 var(--space-sm);display:flex;align-items:center}.c-epPExb{color:var(--colors-actionFgDefault)}.c-cwlcYD{border-width:unset;border-style:unset;cursor:pointer;background-color:var(--colors-actionBgBase);position:relative;border-radius:280px}.c-cwlcYD[data-state="checked"]{background-color:var(--colors-actionBgOn)}.c-hmOgCM{display:block;background-color:var(--colors-actionFgOn);border-radius:9999px;box-shadow:0 1px 3px rgba(0, 0, 0, 0.2);will-change:transform}.c-kYsbuo{position:absolute;outline:3px solid #b6251b;top:1.5625%;left:4.6875%;right:54.6875%;bottom:1.5625%}.c-gRrrJB{position:absolute;outline:3px solid #b6251b;top:1.5625%;left:48.4375%;right:1.5625%;bottom:1.5625%}.c-jEISYs{position:absolute;outline:3px solid #57920d;top:57.8125%;left:39.0625%;right:35.9375%;bottom:1.5625%}.c-cQehLr{display:flex;align-items:baseline;justify-content:flex-start;border-radius:var(--borderRadius-sm);color:var(--colors-feedbackFgReverse);font-size:var(--fontSizes-ms);padding:var(--space-sm)}.c-cQehLr .c-UazGY{margin-left:auto}.c-cQehLr a,.c-cQehLr a:hover,.c-cQehLr .c-cLUmFK,.c-cQehLr .c-cLUmFK:hover{color:inherit}.c-dsPRoj{display:flex;align-items:center;justify-content:center}.c-dsPRoj,.c-dsPRoj > svg{width:var(--space-md);height:var(--space-md)}.c-dsPRoj{margin:0 var(--space-sm) 0;position:relative;top:var(--space-xs)}.c-jgxVxd{display:grid;gap:32px}@media (max-width: 576px){.c-jgxVxd{gap:16px}}}--sxs{--sxs:3 c-gVWXCF-kvkpac-size-large c-gVWXCF-hMocfV-size-small c-dhzjXW-irEjuD-align-stretch c-dhzjXW-ejCoEP-direction-row c-dhzjXW-bICGYT-justify-center c-dhzjXW-kVNAnR-wrap-nowrap c-dhzjXW-iTKOFX-direction-column c-dhzjXW-awKDG-justify-flex-start c-TKGAm-dJzfbb-size-standard c-TKGAm-hGzxBS-tone-tertiary c-iRRrYf-iCTXjZ-showButtons-true c-dyDYjC-ghTPkg-type-left c-dyDYjC-fhLoBm-hide-true c-dyDYjC-lifYVs-type-right c-jUxYmc-beqYnI-isActive-true c-dhzjXW-jroWjL-align-center c-PJLV-foHfor-size-medium c-PJLV-edUNDZ-size-medium c-dhzjXW-XefLA-wrap-wrap c-dhzjXW-lpoiht-gap-md c-dhzjXW-lGOxV-gap-lg c-dhzjXW-bZmKkd-justify-flex-end c-dhzjXW-hNPEVI-gap-xs c-dhzjXW-dQhbEp-gap-sm c-cxpKzu-kqUIUX-size-standard c-dhzjXW-jYFQDJ-gap-ss c-jOMmSw-dAcCEA-size-standard c-dhzjXW-kRIfoH-align-baseline c-dhzjXW-dkuYHy-gap-ms c-fHYEhm-fzEYtA-size-small c-cQcYE-clGKKD-size-standard c-cQcYE-kczqbw-disabled-true c-dhzjXW-igtkwp-align-flex-end c-dhzjXW-gtsvqC-direction-row-reverse c-dhzjXW-knmidH-justify-space-between c-kAArxJ-dZaMxF-variant-small c-kAArxJ-cjiAHf-variant-label c-kAArxJ-jQlRHy-variant-helper c-hakyQ-dPEAfN-spaceY-md c-fyVBxt-kczqbw-disabled-true c-gGIOsy-dncqpF-size-standard c-dhzjXW-ktOuEZ-gap-none c-kAArxJ-gHgXoD-variant-h3 c-kAArxJ-fErRcq-variant-h4 c-epPExb-hgLaUL-size-standard c-cwlcYD-jvbRMk-size-standard c-hmOgCM-jUWmWH-size-standard c-PJLV-dZdqtR-size-large c-PJLV-emaeLH-size-large c-dhzjXW-JrrAq-align-flex-start c-hakyQ-gGUHdm-space-md c-dhzjXW-iBxlCw-gap-ml c-fyVBxt-eSbqjr-status-invalid c-kAArxJ-jIuaGQ-variant-strong c-kAArxJ-iCqGGD-variant-h5 c-cQehLr-kkBrsd-status-info c-dsPRoj-mUjni-status-info c-hakyQ-hrJoyl-spaceY-sm c-jgxVxd-ktOuEZ-gap-none}@media{.c-gVWXCF-kvkpac-size-large{width:600px;padding:var(--space-lg)}.c-gVWXCF-hMocfV-size-small{width:300px;padding:var(--space-md)}.c-dhzjXW-irEjuD-align-stretch{align-items:stretch}.c-dhzjXW-ejCoEP-direction-row{flex-direction:row}.c-dhzjXW-bICGYT-justify-center{justify-content:center}.c-dhzjXW-kVNAnR-wrap-nowrap{flex-wrap:nowrap}.c-dhzjXW-iTKOFX-direction-column{flex-direction:column}.c-dhzjXW-awKDG-justify-flex-start{justify-content:flex-start}.c-TKGAm-dJzfbb-size-standard{font-size:var(--fontSizes-ms);padding:var(--space-sm);border-radius:var(--borderRadius-md)}.c-TKGAm-hGzxBS-tone-tertiary{background-color:transparent;border-color:transparent;color:var(--colors-actionFgTertiary)}.c-TKGAm-hGzxBS-tone-tertiary:hover{background-color:var(--colors-actionBgTertiaryHover);border-color:transparent;color:var(--colors-actionFgTertiaryHover)}.c-TKGAm-hGzxBS-tone-tertiary:active{background-color:var(--colors-actionBgTertiaryActive);border-color:transparent;color:var(--colors-actionFgTertiaryActive)}.c-TKGAm-hGzxBS-tone-tertiary:disabled{background-color:transparent;border-color:transparent;color:var(--colors-actionFgTertiaryDisabled)}.c-iRRrYf-iCTXjZ-showButtons-true{overflow-x:hidden}.c-dyDYjC-ghTPkg-type-left{background:linear-gradient(90deg, var(---gradientColor), transparent);inset:0 auto 0 0}.c-dyDYjC-fhLoBm-hide-true{opacity:0;pointer-events:none}.c-dyDYjC-lifYVs-type-right{background:linear-gradient(90deg, transparent, var(---gradientColor));inset:0 0 0 auto}.c-jUxYmc-beqYnI-isActive-true{font-family:var(--fonts-nvidiaSans);font-weight:var(--fontWeights-medium);line-height:var(--lineHeights-text);font-size:var(--fontSizes-ms);letter-spacing:var(--letterSpacings-none);text-case:none;text-decoration:none;paragraph-spacing:0;color:var(--colors-menuFgSelected)}.c-jUxYmc-beqYnI-isActive-true:before{background-color:var(--colors-menuBorderSelected);height:4px}.c-dhzjXW-jroWjL-align-center{align-items:center}.c-PJLV-foHfor-size-medium{gap:var(--space-md);filter:drop-shadow(0 0 calc(var(--space-md) * 1.5) color-mix(in srgb, var(--colors-feedbackFgProgress), transparent 40%))}.c-PJLV-edUNDZ-size-medium{line-height:0;height:calc(var(--space-md) * 4)}.c-dhzjXW-XefLA-wrap-wrap{flex-wrap:wrap}.c-dhzjXW-lpoiht-gap-md{gap:var(--space-md)}.c-dhzjXW-lGOxV-gap-lg{gap:var(--space-lg)}.c-dhzjXW-bZmKkd-justify-flex-end{justify-content:flex-end}.c-dhzjXW-hNPEVI-gap-xs{gap:var(--space-xs)}.c-dhzjXW-dQhbEp-gap-sm{gap:var(--space-sm)}.c-cxpKzu-kqUIUX-size-standard{font-size:var(--fontSizes-ms);height:var(--space-lg);padding:var(--space-sm)}.c-dhzjXW-jYFQDJ-gap-ss{gap:var(--space-ss)}.c-jOMmSw-dAcCEA-size-standard{font-size:var(--fontSizes-ms);padding:var(--space-ss) var(--space-sm)}.c-dhzjXW-kRIfoH-align-baseline{align-items:baseline}.c-dhzjXW-dkuYHy-gap-ms{gap:var(--space-ms)}.c-fHYEhm-fzEYtA-size-small{height:var(--space-md);width:var(--space-md)}.c-cQcYE-clGKKD-size-standard{font-size:var(--fontSizes-ms);height:var(--space-lg);padding-left:var(--space-sm);padding-right:var(--space-sm)}.c-cQcYE-kczqbw-disabled-true{background-color:var(--colors-inputBgDisabled);border-color:var(--colors-inputBorderDisabled);color:var(--colors-inputFgDisabled);cursor:not-allowed}.c-cQcYE-kczqbw-disabled-true:hover{border-color:var(--colors-inputBorderDisabled)}.c-dhzjXW-igtkwp-align-flex-end{align-items:flex-end}.c-dhzjXW-gtsvqC-direction-row-reverse{flex-direction:row-reverse}.c-dhzjXW-knmidH-justify-space-between{justify-content:space-between}.c-kAArxJ-dZaMxF-variant-small{---fontSize:var(--fontSizes-ms)}.c-kAArxJ-cjiAHf-variant-label{---fontSize:var(--fontSizes-sm);---fontWeight:var(--fontWeights-medium);---lineHeight:var(--lineHeights-text);color:var(--colors-inputFgLabel)}.c-kAArxJ-jQlRHy-variant-helper{---fontSize:var(--fontSizes-sm);---fontWeight:var(--fontWeights-regular);---lineHeight:var(--lineHeights-text);color:var(--colors-inputFgDefault)}.c-hakyQ-dPEAfN-spaceY-md{padding-bottom:var(--space-md);padding-top:var(--space-md)}.c-fyVBxt-kczqbw-disabled-true{background-color:var(--colors-inputBgDisabled);border-color:var(--colors-inputBorderDisabled);color:var(--colors-inputFgDisabled);cursor:not-allowed}.c-fyVBxt-kczqbw-disabled-true:hover{border-color:var(--colors-inputBorderDisabled)}.c-gGIOsy-dncqpF-size-standard{font-size:var(--fontSizes-lg);font-weight:var(--fontWeights-medium);line-height:var(--lineHeights-heading)}.c-dhzjXW-ktOuEZ-gap-none{gap:var(--space-none)}.c-kAArxJ-gHgXoD-variant-h3{---fontSize:var(--fontSizes-ml);---fontWeight:var(--fontWeights-medium);---lineHeight:var(--lineHeights-heading)}.c-kAArxJ-fErRcq-variant-h4{---fontSize:var(--fontSizes-md);---fontWeight:var(--fontWeights-medium);---lineHeight:var(--lineHeights-heading)}.c-epPExb-hgLaUL-size-standard{font-size:var(--fontSizes-ms)}.c-cwlcYD-jvbRMk-size-standard{width:var(--space-lg);height:var(--space-md)}.c-hmOgCM-jUWmWH-size-standard{width:var(--space-ms);height:var(--space-ms);transform:translateX(2px)}.c-hmOgCM-jUWmWH-size-standard[data-state="checked"]{transform:translateX(18px)}.c-PJLV-dZdqtR-size-large{gap:var(--space-lg)}.c-PJLV-dZdqtR-size-large svg{filter:drop-shadow(0 0 calc(var(--space-lg) * .75) color-mix(in srgb, var(--colors-feedbackFgProgress), transparent 40%))}.c-PJLV-emaeLH-size-large{line-height:0;height:calc(var(--space-lg) * 4)}.c-dhzjXW-JrrAq-align-flex-start{align-items:flex-start}.c-hakyQ-gGUHdm-space-md{padding:var(--space-md)}.c-dhzjXW-iBxlCw-gap-ml{gap:var(--space-ml)}.c-fyVBxt-eSbqjr-status-invalid{border-image:linear-gradient(to top, var(--colors-inputBorderError), var(--colors-inputBorderError) 2px, transparent 2px, transparent) 2}.c-fyVBxt-eSbqjr-status-invalid:hover{border-image:linear-gradient(to top, var(--colors-inputBorderError), var(--colors-inputBorderError) 2px, transparent 2px, transparent) 2}.c-fyVBxt-eSbqjr-status-invalid:focus-within{border-image:linear-gradient(to top, var(--colors-inputBorderError), var(--colors-inputBorderError) 2px, transparent 2px, transparent) 2}.c-kAArxJ-jIuaGQ-variant-strong{---fontWeight:var(--fontWeights-bold)}.c-kAArxJ-iCqGGD-variant-h5{---fontSize:var(--fontSizes-ms);---fontWeight:var(--fontWeights-medium);---lineHeight:var(--lineHeights-heading)}.c-cQehLr-kkBrsd-status-info{background:linear-gradient(to right, var(--colors-feedbackBorderInfo), var(--colors-feedbackBorderInfo) var(--borderWidths-lg), var(--colors-feedbackBgInfo) var(--borderWidths-lg), var(--colors-feedbackBgInfo))}.c-dsPRoj-mUjni-status-info{color:var(--colors-feedbackBorderInfo)}.c-hakyQ-hrJoyl-spaceY-sm{padding-bottom:var(--space-sm);padding-top:var(--space-sm)}.c-jgxVxd-ktOuEZ-gap-none{gap:var(--space-none)}}--sxs{--sxs:6 c-dwAUIm-iORGnN-css c-coDgbF-iPJLV-css c-TKGAm-ikBYYkb-css c-dyDYjC-ictjieU-css c-jUxYmc-ieaTcvE-css c-jUxYmc-ikXXknH-css c-dhzjXW-icmpvrW-css c-dhzjXW-ikdltkg-css c-Dadvv-icMSEqv-css c-Dadvv-icbWVQp-css c-Dadvv-ijEjRjM-css c-Dadvv-idLADTn-css c-Dadvv-icXGNHg-css c-Dadvv-ibDfrWj-css c-Dadvv-iQNGUs-css c-Dadvv-ihboaFE-css c-Dadvv-iflEUYh-css c-Dadvv-iklerJH-css c-Dadvv-ijtizoc-css c-Dadvv-ifpBmpE-css c-Dadvv-ibtowYx-css c-Dadvv-ihVERPj-css c-Dadvv-ijOoRKQ-css c-dhzjXW-ifGHEql-css c-hakyQ-ikxhBIE-css c-fpDrvI-icmecNs-css c-cxpKzu-igSRhde-css c-PJLV-ifxrEBZ-css c-dhzjXW-ijhxVpd-css c-dhzjXW-ijcpczY-css c-dhzjXW-ieiLNWk-css c-jOMmSw-ibvlxRt-css c-dhzjXW-iYNHLz-css c-fpDrvI-idqQkoG-css c-fpDrvI-ijeDiRL-css c-dhzjXW-iiCZUBl-css c-hakyQ-ibPNjhd-css c-duhgWy-iPJLV-css c-dWeQby-iPJLV-css c-dCzjnn-igcOMTV-css c-fHYEhm-iPJLV-css c-ijtrue-ibFfFKv-css c-kMaXzg-iPJLV-css c-eFiwTo-iPJLV-css c-dhzjXW-idvXSds-css c-cQcYE-ijaQKCi-css c-kzncD-ilcpctE-css c-dhzjXW-iknpVsn-css c-dhzjXW-iYKtYl-css c-hakyQ-ibxUnUc-css c-cQcYE-ihFdqvS-css c-dhzjXW-ihiYLFF-css c-dhzjXW-ilkBNdM-css c-dMjTYf-ihHQxct-css c-kAArxJ-iiyTxCE-css c-kAArxJ-ibGvNaJ-css c-dhzjXW-ieamsnf-css c-dhzjXW-iiNirjJ-css c-dhzjXW-ikYXVsi-css c-hakyQ-ieGaVeY-css c-dhzjXW-ijaeFnC-css c-kAArxJ-ifOCFMN-css c-hakyQ-ihWMCax-css c-dhzjXW-ilkiHBK-css c-dhzjXW-icAnTdq-css c-dhzjXW-ikAQipp-css c-kAArxJ-icVcmBF-css c-gslCyW-icDfPFC-css c-ElNdM-icBQnZI-css c-MCGcG-idDnJxY-css c-jTLszv-icCIxyn-css c-dhzjXW-iPJLV-css c-hakyQ-ieQbzoM-css c-hakyQ-iiOfNaz-css c-dhzjXW-ifijhlb-css c-dhzjXW-icxaDwp-css c-cQcYE-iiPuBnJ-css c-cxpKzu-ieJGvjC-css c-dhzjXW-ihZXyPW-css c-cxpKzu-iLkYCP-css c-dhzjXW-ieBEpVQ-css c-hakyQ-igPhOUD-css c-dhzjXW-iecTQhR-css c-kAArxJ-idbzzBd-css c-kAArxJ-ibitErx-css c-dhzjXW-ibPNjhd-css c-dhzjXW-ibHUWRq-css c-dhzjXW-illJumK-css c-dhzjXW-ifMQoSC-css c-dhzjXW-ieaPeMQ-css c-dhzjXW-ifiJtrg-css c-dhzjXW-ikwxibr-css c-hakyQ-iiFJMwr-css c-dhzjXW-ijXkSAP-css c-fpDrvI-iiqYbsL-css c-kAArxJ-ijqdZGR-css c-dhzjXW-ikQAsCx-css c-jOMmSw-iklIoOk-css c-hakyQ-iiYDnlq-css c-gslCyW-iblfBiT-css c-dhzjXW-irAPDU-css c-dhzjXW-ikmKlsb-css c-hakyQ-ifHcIka-css c-kAArxJ-ijXxGIw-css c-dhzjXW-ieFNXWk-css c-iGJBGr-igbMWsE-css c-epPExb-igbMWsE-css c-cwlcYD-igbMWsE-css c-hmOgCM-igbMWsE-css c-hakyQ-ifCWwyj-css c-dhzjXW-igoNsge-css c-hakyQ-ieWVJxx-css c-dhzjXW-ikDpbbZ-css c-kAArxJ-igWpoIe-css c-dhzjXW-icOXzkM-css c-dhzjXW-iymzpH-css c-gslCyW-ifGHEql-css c-hakyQ-ibhyizs-css c-hakyQ-ijHyYJs-css c-kAArxJ-ifvDikr-css c-hakyQ-ikVSwXF-css c-kAArxJ-idbuuRq-css c-kAArxJ-ibFeFXz-css c-dhzjXW-ienKwaM-css c-hakyQ-icNkccz-css c-cxpKzu-ikquKKs-css c-dhzjXW-ijHycBS-css c-fpDrvI-iixrZis-css c-iGJBGr-iAfGps-css c-epPExb-iAfGps-css c-cwlcYD-iAfGps-css c-hmOgCM-iAfGps-css c-dhzjXW-ihvAWJs-css c-cxpKzu-ikUDIcY-css c-hakyQ-ierkLVf-css c-hakyQ-ikiIXZF-css c-hakyQ-ikqZLqP-css c-dhzjXW-ikGSFxO-css c-cxpKzu-ijhVimD-css c-dhzjXW-ibLUdNl-css c-kAArxJ-iSMsfV-css c-dhzjXW-ikAnjbF-css c-dhzjXW-ijewkys-css c-hakyQ-icCXlWf-css c-dhzjXW-ieGaVeY-css c-dhzjXW-igEIieZ-css c-jgxVxd-ilgQfFT-css c-jgxVxd-ihbthGk-css c-jgxVxd-idrAaWl-css c-dhzjXW-ikdyFAG-css c-dhzjXW-iculkRU-css c-hakyQ-idchnVT-css c-kAArxJ-ifEHJXU-css c-hakyQ-iiHKkUW-css c-hakyQ-ikGTPTu-css c-hakyQ-ijVDGcn-css c-hakyQ-ibrsHGX-css c-hakyQ-icInFh-css c-hakyQ-ihugmgO-css c-hakyQ-idJRZry-css c-hakyQ-ikYDmCv-css}@media{.c-dwAUIm-iORGnN-css{--colors-displayBgCanvas:transparent}.c-coDgbF-iPJLV-css{max-height:}.c-TKGAm-ikBYYkb-css{display:none}.c-dyDYjC-ictjieU-css{---gradientColor:var(--colors-menuBgBase)}.c-jUxYmc-ieaTcvE-css{line-height:var(--lineHeights-body);color:#76b900;font-weight:var(--fontWeights-medium)}.c-jUxYmc-ieaTcvE-css:before{border-radius:10px;background:linear-gradient(to left, #6EA60C, #BBD888)}.c-jUxYmc-ikXXknH-css{line-height:var(--lineHeights-body)}.c-jUxYmc-ikXXknH-css color{-d-e-f-a-u-l-t:#919191;active:#76b900}.c-jUxYmc-ikXXknH-css color secondary{-d-e-f-a-u-l-t:#787878;active:#E0E0E0}.c-jUxYmc-ikXXknH-css{font-weight:var(--fontWeights-medium)}.c-jUxYmc-ikXXknH-css:before{border-radius:10px;background:transparent}.c-dhzjXW-icmpvrW-css{position:relative}.c-dhzjXW-ikdltkg-css{min-height:200px}.c-Dadvv-icMSEqv-css{animation-delay:-1s}.c-Dadvv-icbWVQp-css{animation-delay:-.9444s}.c-Dadvv-ijEjRjM-css{animation-delay:-.8888s}.c-Dadvv-idLADTn-css{animation-delay:-.8333s}.c-Dadvv-icXGNHg-css{animation-delay:-.7777s}.c-Dadvv-ibDfrWj-css{animation-delay:0s}.c-Dadvv-iQNGUs-css{animation-delay:-.1111s}.c-Dadvv-ihboaFE-css{animation-delay:-.6666s}.c-Dadvv-iflEUYh-css{animation-delay:-.7222s}.c-Dadvv-iklerJH-css{animation-delay:-.1666s}.c-Dadvv-ijtizoc-css{animation-delay:-.2222s}.c-Dadvv-ifpBmpE-css{animation-delay:-.3333s}.c-Dadvv-ibtowYx-css{animation-delay:-.5s}.c-Dadvv-ihVERPj-css{animation-delay:-.6111s}.c-Dadvv-ijOoRKQ-css{animation-delay:-.5555s}.c-dhzjXW-ifGHEql-css{width:100%}.c-hakyQ-ikxhBIE-css{width:100%;border:1px solid #5E5E5E !important;padding:var(--space-md) var(--space-lg);border-radius:8px}.c-fpDrvI-icmecNs-css{font-size:var(--fontSizes-md)}.c-cxpKzu-igSRhde-css{background:#121212;border-color:#5E5E5E;border-style:solid;border-width:var(--borderWidths-sm) !important;border-radius:8px;border-image:none !important;color:#f2f2f2;font-size:14px;height:unset;padding:var(--space-md) var(--space-lg);border:var(--borderWidths-sm) solid #5E5E5E}.c-cxpKzu-igSRhde-css[data-state="open"]{background:#121212;border:}.c-cxpKzu-igSRhde-css[data-placeholder]{color:#767676;font-family:var(--fonts-nvidiaSans)}.c-cxpKzu-igSRhde-css:hover{cursor:pointer;border:var(--borderWidths-sm) solid #8f8f8f}.c-cxpKzu-igSRhde-css{& .select-item-container svg:}.c-PJLV-ifxrEBZ-css{overflow:hidden}.c-dhzjXW-ijhxVpd-css{word-wrap:break-word}.c-dhzjXW-ijcpczY-css{flex:1;border:1px solid #5E5E5E !important;background-color:#121212;border-radius:8px}.c-dhzjXW-ijcpczY-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-ijcpczY-css:focus-within{border-color:#76b900 !important;background-color:#121212 !important}.c-dhzjXW-ieiLNWk-css{padding:var(--space-md) var(--space-lg);flex:1}.c-jOMmSw-ibvlxRt-css{color:#f2f2f2;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);line-height:var(--lineHeights-body);padding:0 !important;pointer-events:auto}.c-dhzjXW-iYNHLz-css{margin-top:16px;margin-bottom:8px;display:flex;flex-direction:column;gap:8px}.c-fpDrvI-idqQkoG-css{margin-bottom:8px}.c-fpDrvI-ijeDiRL-css{color:#FFFFFF;margin-top:16px;margin-bottom:8px;display:flex;align-items:center;gap:8px}.c-dhzjXW-iiCZUBl-css{padding-top:var(--space-sm);padding-bottom:var(--space-md);width:100%}.c-hakyQ-ibPNjhd-css{flex:1}.c-dCzjnn-igcOMTV-css{background-color:unset}.c-ijtrue-ibFfFKv-css{margin-top:0 !important;width:unset !important}.c-dhzjXW-idvXSds-css [data-testid="kui-text-input-container"]{padding:0}.c-dhzjXW-idvXSds-css input{text-align:center;padding:var(--space-md) var(--space-sm);width:60px;height:}.c-cQcYE-ijaQKCi-css{background-color:#121212;color:;border:1px solid #5E5E5E !important;border-radius:8px;height:unset;margin:0 !important;padding-left:var(--space-md);padding-right:var(--space-md)}.c-cQcYE-ijaQKCi-css:hover{border-color:#8f8f8f !important}.c-cQcYE-ijaQKCi-css:focus-within{border-color:#76b900 !important;background-color:#202020 !important;color:#f2f2f2 !important}.c-kzncD-ilcpctE-css{flex:1;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);line-height:var(--lineHeights-body);padding-top:var(--space-ms);padding-bottom:var(--space-ms)}.c-dhzjXW-iknpVsn-css{margin-top:32px}.c-dhzjXW-iYKtYl-css{margin-top:16px}.c-hakyQ-ibxUnUc-css{width:100%;border:1px solid #5E5E5E !important;padding:var(--space-md) var(--space-lg);border-radius:8px;color:#f2f2f2}.c-cQcYE-ihFdqvS-css{background-color:#202020;color:#767676;border:1px solid #5E5E5E !important;border-radius:8px;height:unset;margin:0 !important;padding-left:var(--space-md);padding-right:var(--space-md)}.c-cQcYE-ihFdqvS-css:hover{border-color:#8f8f8f !important}.c-cQcYE-ihFdqvS-css:focus-within{border-color:#76b900 !important;background-color:#202020 !important;color:#f2f2f2 !important}.c-dhzjXW-ihiYLFF-css{flex:1;padding:var(--space-lg);padding-right:0;overflow:hidden}.c-dhzjXW-ilkBNdM-css{height:100%}.c-dMjTYf-ihHQxct-css div[role="tablist"]{padding-bottom:var(--space-sm)}.c-dMjTYf-ihHQxct-css div[role="tablist"]::before{height:1px}.c-dMjTYf-ihHQxct-css div[data-testid="kui-tab-list"] > div:first-child{overflow-x:hidden}.c-dMjTYf-ihHQxct-css{padding-bottom:var(--space-md)}.c-kAArxJ-iiyTxCE-css{font-weight:500;font-size:24px;line-height:28.8px;color:#F2F2F2}.c-kAArxJ-ibGvNaJ-css{font-size:14px;line-height:21px;color:#C6C6C6}.c-dhzjXW-ieamsnf-css{background-color:#121212;border:1px solid #5E5E5E !important;border-radius:8px;color:#f2f2f2;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);height:unset;line-height:var(--lineHeights-body);padding:none;position:relative}.c-dhzjXW-iiNirjJ-css{background-image:linear-gradient(180deg, #121212 80%, rgba(16,16,16,0) 100%);position:absolute;width:100%;border-radius:8px;padding-top:var(--space-ms);padding-bottom:var(--space-ms);padding-left:var(--space-md);padding-right:var(--space-md);z-index:10}.c-dhzjXW-ikYXVsi-css{margin-bottom:var(--space-sm);overflow:auto;background-color:#121212;border:none;border-radius:8px;color:#f2f2f2;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);height:400px;line-height:var(--lineHeights-body);padding:var(--space-md) var(--space-lg);margin-top:40px}@media screen and (max-width: 639px){.c-dhzjXW-ikYXVsi-css{padding:var(--space-sm)}}.c-hakyQ-ieGaVeY-css{flex-grow:1}@media screen and (max-width: 639px){.c-dhzjXW-ijaeFnC-css{margin-right:var(--space-sm)}}.c-kAArxJ-ifOCFMN-css{font-size:var(--fontSizes-ms);font-family:var(--fonts-nvidiaSans);margin-bottom:var(--space-md)}.c-hakyQ-ihWMCax-css{pointer-events:none}.c-dhzjXW-ilkiHBK-css{border:1px solid #5E5E5E !important;background-color:#121212 !important;border-radius:8px !important;width:100%}.c-dhzjXW-ilkiHBK-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-ilkiHBK-css:focus-within{border-color:#76b900 !important}.c-dhzjXW-icAnTdq-css{padding-right:0;width:100%}.c-dhzjXW-ikAQipp-css{flex:1;border:none !important;background-color:#121212;border-radius:8px}.c-dhzjXW-ikAQipp-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-ikAQipp-css:focus-within{border-color:#76b900 !important;background-color:#121212 !important}.c-kAArxJ-icVcmBF-css{align-self:flex-end;display:none}.c-gslCyW-icDfPFC-css{margin-top:var(--space-lg)}.c-ElNdM-icBQnZI-css{background:#121212}.c-ElNdM-icBQnZI-css div[data-state='open']{overflow:hidden;animation:k-gLaGBR 250ms ease-out}.c-ElNdM-icBQnZI-css div[data-state='closed']{overflow:hidden;animation:k-jBKOzt 250ms ease-out}.c-MCGcG-idDnJxY-css:disabled span{color:#8f8f8f}.c-MCGcG-idDnJxY-css:hover:not(:disabled){background-color:#202020}.c-MCGcG-idDnJxY-css{color:#ffffff;gap:var(--space-ms);width:auto;flex:1;padding:20px var(--space-md)}.c-jTLszv-icCIxyn-css > span{color:#E0E0E0;font-weight:var(--fontWeights-bold);font-size:var(--fontSizes-ms)}.c-hakyQ-ieQbzoM-css{border:1px solid #5E5E5E !important;padding:var(--space-md) var(--space-lg);border-radius:8px;font-size:14px;color:#f2f2f2;position:relative;max-height:100%;overflow:auto;background-color:#121212;flex:1;padding-top:0;margin-top:0}.c-hakyQ-iiOfNaz-css{position:relative}.c-hakyQ-iiOfNaz-css.c-hakyQ-iiOfNaz-css pre{white-space:pre-wrap;word-wrap:break-word}.c-hakyQ-iiOfNaz-css code[class*='language-'],.c-hakyQ-iiOfNaz-css pre[class*='language-']{word-break:break-all !important}.c-hakyQ-iiOfNaz-css{overflow-y:auto;font-family:var(--fonts-nvidiaSans);max-height:Infinitypx;margin-top:var(--space-md)}.c-hakyQ-iiOfNaz-css span.string{word-break:break-all}.c-dhzjXW-ifijhlb-css{background-image:linear-gradient(180deg, #121212 80%, rgba(16,16,16,0) 100%)}.c-dhzjXW-icxaDwp-css svg{display:none}@media (max-width: 576px){.c-dhzjXW-icxaDwp-css svg{display:block !important}}.c-cQcYE-iiPuBnJ-css{background-color:#121212;color:;border:1px solid #5E5E5E !important;border-radius:45px;height:unset;margin:0 !important;padding-left:var(--space-md);padding-right:var(--space-md)}.c-cQcYE-iiPuBnJ-css:hover{border-color:#8f8f8f !important}.c-cQcYE-iiPuBnJ-css:focus-within{border-color:#76b900 !important;background-color:#202020 !important;color:#f2f2f2 !important}.c-cQcYE-iiPuBnJ-css > input::placeholder{color:#767676}.c-cxpKzu-ieJGvjC-css{background:#121212;border-color:#5E5E5E;border-style:solid;border-width:var(--borderWidths-sm) !important;border-radius:45px;border-image:none !important;color:#f2f2f2;font-size:14px;height:45px;padding:var(--space-md) var(--space-md);border:var(--borderWidths-sm) solid #5E5E5E}.c-cxpKzu-ieJGvjC-css[data-state="open"]{background:#121212;border:var(--borderWidths-sm) solid #8fcb2b}.c-cxpKzu-ieJGvjC-css[data-state="open"]::before{color:#f2f2f2}.c-cxpKzu-ieJGvjC-css[data-state="open"] div > span{color:#f2f2f2}.c-cxpKzu-ieJGvjC-css[data-placeholder]{color:#767676;font-family:var(--fonts-nvidiaSans)}.c-cxpKzu-ieJGvjC-css:hover{cursor:pointer;border:var(--borderWidths-sm) solid #8f8f8f}.c-cxpKzu-ieJGvjC-css .select-item-container svg{display:none}.c-cxpKzu-ieJGvjC-css::before{content:"Sort By:";padding-right:5px;white-space:nowrap;color:#767676}.c-cxpKzu-ieJGvjC-css div:first-child > span{margin-right:var(--space-sm)}.c-cxpKzu-ieJGvjC-css div > span{color:#767676}.c-dhzjXW-ihZXyPW-css{width:175px}.c-cxpKzu-iLkYCP-css{color:var(--colors-inputFgDefault)}.c-dhzjXW-ieBEpVQ-css{margin-right:var(--space-xs)}.c-hakyQ-igPhOUD-css{position:absolute;bottom:-25px;height:25px;width:100%;background:linear-gradient(to bottom, rgba(12, 12, 12, 1) 0%, rgba(12, 12, 12, 0) 100%)}.c-dhzjXW-iecTQhR-css.c-dhzjXW-iecTQhR-css{justify-content:flex-end}.c-kAArxJ-idbzzBd-css{white-space:nowrap}.c-kAArxJ-ibitErx-css{line-height:var(--lineHeights-heading);text-align:center}.c-dhzjXW-ibPNjhd-css{flex:1}.c-dhzjXW-ibHUWRq-css{position:relative;border:1px solid #5E5E5E !important;border-radius:8px;overflow:hidden}.c-dhzjXW-illJumK-css{height:310px}.c-dhzjXW-ifMQoSC-css{padding:var(--space-sm);color:#f2f2f2}.c-dhzjXW-ieaPeMQ-css{display:none}@media (max-width: 576px){.c-dhzjXW-ieaPeMQ-css{display:flex;align-self:flex-end}}.c-dhzjXW-ifiJtrg-css{background-color:#121212}.c-dhzjXW-ikwxibr-css{border:1px solid #5E5E5E !important;border-radius:8px;height:380px;padding:}.c-hakyQ-iiFJMwr-css{text-overflow:ellipsis;white-wpace:nowrap;overflow:hidden}.c-dhzjXW-ijXkSAP-css{color:#f2f2f2;font-size:var(--fontSizes-sm)}.c-fpDrvI-iiqYbsL-css{font-size:var(--fontSizes-lg)}.c-kAArxJ-ijqdZGR-css{min-height:12px;color:#f2f2f2}.c-dhzjXW-ikQAsCx-css{flex:1;border:1px solid #5E5E5E !important;background-color:#202020;border-radius:8px}.c-dhzjXW-ikQAsCx-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-ikQAsCx-css:focus-within{border-color:#76b900 !important;background-color:#121212 !important}.c-jOMmSw-iklIoOk-css{color:#767676;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);line-height:var(--lineHeights-body);padding:0 !important;pointer-events:auto}.c-hakyQ-iiYDnlq-css{margin-top:var(--space-lg);width:100%}.c-gslCyW-iblfBiT-css{color:#ffffff;font-family:var(--fonts-nvidiaSans)}.c-gslCyW-iblfBiT-css > div{border-bottom:1px solid #2B2B2B;border-top:1px solid #2B2B2B}.c-gslCyW-iblfBiT-css{display:flex;flex-direction:column;gap:var(--space-ms);background:transparent}.c-dhzjXW-irAPDU-css{padding-bottom:var(--space-xs);width:100%}.c-dhzjXW-ikmKlsb-css{align-self:flex-end;width:100%}.c-hakyQ-ifHcIka-css{background-color:#121212;border:1px solid #5E5E5E !important;border-radius:8px;color:#f2f2f2;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);height:unset;line-height:var(--lineHeights-body);padding:0;overflow:hidden}.c-kAArxJ-ijXxGIw-css{padding-bottom:var(--space-lg);text-align:center}.c-dhzjXW-ieFNXWk-css{flex:1;overflow:auto}.c-iGJBGr-igbMWsE-css{gap:var(--space-ms)}.c-iGJBGr-igbMWsE-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-iGJBGr-igbMWsE-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-iGJBGr-igbMWsE-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-iGJBGr-igbMWsE-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-iGJBGr-igbMWsE-css{padding-right:var(--space-xxs)}.c-epPExb-igbMWsE-css{gap:var(--space-ms)}.c-epPExb-igbMWsE-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-epPExb-igbMWsE-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-epPExb-igbMWsE-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-epPExb-igbMWsE-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-epPExb-igbMWsE-css{padding-right:var(--space-xxs)}.c-cwlcYD-igbMWsE-css{gap:var(--space-ms)}.c-cwlcYD-igbMWsE-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-cwlcYD-igbMWsE-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-cwlcYD-igbMWsE-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-cwlcYD-igbMWsE-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-cwlcYD-igbMWsE-css{padding-right:var(--space-xxs)}.c-hmOgCM-igbMWsE-css{gap:var(--space-ms)}.c-hmOgCM-igbMWsE-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-hmOgCM-igbMWsE-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-hmOgCM-igbMWsE-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-hmOgCM-igbMWsE-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-hmOgCM-igbMWsE-css{padding-right:var(--space-xxs)}.c-hakyQ-ifCWwyj-css{max-height:350px;overflow:auto}.c-dhzjXW-igoNsge-css{margin-top:var(--space-md)}.c-hakyQ-ieWVJxx-css{position:relative;width:100%;height:100%;min-height:500px}.c-dhzjXW-ikDpbbZ-css{height:405px}.c-kAArxJ-igWpoIe-css{font-weight:var(--fontWeights-regular);font-size:var(--fontSizes-sm);margin-bottom:var(--space-xxs)}.c-dhzjXW-icOXzkM-css{width:100%;min-height:600px}.c-dhzjXW-iymzpH-css{padding:var(--space-md);border:1px solid #5E5E5E;border-radius:20px}.c-gslCyW-ifGHEql-css{width:100%}.c-hakyQ-ibhyizs-css{border:1px solid #5E5E5E !important;border-radius:8px;width:100%;max-height:700px;display:flex;align-items:center;justify-content:center}.c-hakyQ-ijHyYJs-css{border-radius:8px;border:1px solid #5E5E5E !important;width:100%;min-height:100px}.c-hakyQ-ijHyYJs-css .bounding-box-phrase{border-radius:5px;padding:var(--space-xs);line-height:200%}.c-kAArxJ-ifvDikr-css{line-height:var(--lineHeights-body)}.c-hakyQ-ikVSwXF-css{padding-left:var(--space-md);flex:1}.c-kAArxJ-idbuuRq-css{color:#c6c6c6;font-weight:var(--fontWeights-regular);font-size:var(--fontSizes-sm);margin-bottom:var(--space-xxs)}.c-kAArxJ-idbuuRq-css ~ span{border-left:1px solid #767676;padding-left:var(--space-ss)}.c-kAArxJ-ibFeFXz-css{align-self:flex-end}.c-dhzjXW-ienKwaM-css{display:inherit}.c-hakyQ-icNkccz-css{width:100%;height:30px}.c-cxpKzu-ikquKKs-css{background:#121212;border-color:#5E5E5E;border-style:solid;border-width:var(--borderWidths-sm) !important;border-radius:8px;border-image:none !important;color:#f2f2f2;font-size:14px;height:unset;padding:var(--space-md);border:var(--borderWidths-sm) solid #5E5E5E}.c-cxpKzu-ikquKKs-css[data-state="open"]{background:#121212;border:}.c-cxpKzu-ikquKKs-css[data-placeholder]{color:#767676;font-family:var(--fonts-nvidiaSans)}.c-cxpKzu-ikquKKs-css:hover{cursor:pointer;border:var(--borderWidths-sm) solid #8f8f8f}.c-cxpKzu-ikquKKs-css{& .select-item-container svg:}.c-dhzjXW-ijHycBS-css{padding:var(--space-md);flex:1}.c-fpDrvI-iixrZis-css{font-family:var(--fonts-jetbrainsMono);background-color:#2C2C2C;border-radius:5px;font-size:var(--fontSizes-sm);font-weight:var(--fontWeights-medium);line-height:var(--lineHeights-body);padding:var(--space-xs) var(--space-ms);text-align:center;text-transform:none}.c-iGJBGr-iAfGps-css{gap:var(--space-ms)}.c-iGJBGr-iAfGps-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-iGJBGr-iAfGps-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-iGJBGr-iAfGps-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-iGJBGr-iAfGps-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-epPExb-iAfGps-css{gap:var(--space-ms)}.c-epPExb-iAfGps-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-epPExb-iAfGps-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-epPExb-iAfGps-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-epPExb-iAfGps-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-cwlcYD-iAfGps-css{gap:var(--space-ms)}.c-cwlcYD-iAfGps-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-cwlcYD-iAfGps-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-cwlcYD-iAfGps-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-cwlcYD-iAfGps-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-hmOgCM-iAfGps-css{gap:var(--space-ms)}.c-hmOgCM-iAfGps-css label{color:#9E9E9E;font-size:var(--fontSizes-sm)}.c-hmOgCM-iAfGps-css button{background:#434343 !important;align-items:center;display:flex;width:22px;height:10px}.c-hmOgCM-iAfGps-css button > span{transform:translateX(-1px);background-color:#8F8F8F}.c-hmOgCM-iAfGps-css button > span[data-state="checked"]{transform:translateX(11px);background-color:#76b900}.c-dhzjXW-ihvAWJs-css{background:#121212;border:1px solid #5E5E5E !important;border-radius:8px}.c-dhzjXW-ihvAWJs-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-ihvAWJs-css{color:#f2f2f2;font-size:var(--fontSizes-ms);height:unset;padding:var(--space-md) var(--space-lg)}.c-dhzjXW-ihvAWJs-css > div{flex:0 0 auto}.c-cxpKzu-ikUDIcY-css{background:#121212;border-color:#5E5E5E;border-style:solid;border-width:var(--borderWidths-sm) !important;border-radius:8px;border-image:none !important;color:#f2f2f2;font-size:14px;height:32px;padding:var(--space-md);border:var(--borderWidths-sm) solid #5E5E5E}.c-cxpKzu-ikUDIcY-css[data-state="open"]{background:#121212;border:}.c-cxpKzu-ikUDIcY-css[data-placeholder]{color:#767676;font-family:var(--fonts-nvidiaSans)}.c-cxpKzu-ikUDIcY-css:hover{cursor:pointer;border:var(--borderWidths-sm) solid #8f8f8f}.c-cxpKzu-ikUDIcY-css{& .select-item-container svg:}.c-hakyQ-ierkLVf-css{border:1px solid #5E5E5E !important;padding:var(--space-md) var(--space-lg);border-radius:8px;font-size:14px;color:#f2f2f2;position:relative;max-height:100%;overflow:auto;background-color:#121212;flex:1}.c-hakyQ-ikiIXZF-css{position:relative}.c-hakyQ-ikiIXZF-css.c-hakyQ-ikiIXZF-css pre{white-space:pre-wrap;word-wrap:unset}.c-hakyQ-ikiIXZF-css code[class*='language-'],.c-hakyQ-ikiIXZF-css pre[class*='language-']{word-break:unset}.c-hakyQ-ikiIXZF-css{overflow-y:auto;font-family:var(--fonts-nvidiaSans);max-height:800px;margin-top:unset}.c-hakyQ-ikiIXZF-css span.string{word-break:break-all}.c-hakyQ-ikqZLqP-css td:not(:first-child){padding-left:var(--space-xxl)}.c-hakyQ-ikqZLqP-css th:not(:first-child){padding-left:var(--space-xxl)}.c-hakyQ-ikqZLqP-css th:first-child{display:flex}.c-dhzjXW-ikGSFxO-css{color:#f2f2f2;font-size:var(--fontSizes-ms);padding:0 0 0 var(--space-sm)}.c-dhzjXW-ikGSFxO-css > div{flex:0 0 auto}.c-cxpKzu-ijhVimD-css{background:#121212;border-color:#5E5E5E;border-style:solid;border-width:var(--borderWidths-sm) !important;border-radius:8px;border-image:none !important;color:#f2f2f2;font-size:14px;height:40px;padding:var(--space-md) var(--space-lg);border:var(--borderWidths-sm) solid #5E5E5E}.c-cxpKzu-ijhVimD-css[data-state="open"]{background:#121212;border:}.c-cxpKzu-ijhVimD-css[data-placeholder]{color:#767676;font-family:var(--fonts-nvidiaSans)}.c-cxpKzu-ijhVimD-css:hover{cursor:pointer;border:var(--borderWidths-sm) solid #8f8f8f}.c-cxpKzu-ijhVimD-css{& .select-item-container svg:}.c-dhzjXW-ibLUdNl-css{margin-bottom:var(--space-sm);overflow:auto;background-color:#121212;border:none;border-radius:8px;color:#f2f2f2;font-family:var(--fonts-nvidiaSans);font-size:var(--fontSizes-ms);height:425px;line-height:var(--lineHeights-body);padding:var(--space-md) var(--space-lg);margin-top:40px}@media screen and (max-width: 639px){.c-dhzjXW-ibLUdNl-css{padding:var(--space-sm)}}.c-kAArxJ-iSMsfV-css{align-self:center;flex:1}.c-dhzjXW-ikAnjbF-css{padding-left:var(--space-md)}@media only screen and (min-width: 1024px){.c-dhzjXW-ikAnjbF-css{max-width:475px;min-width:475px}}@media screen and (max-width: 639px){.c-dhzjXW-ijewkys-css{flex-direction:column}}.c-hakyQ-icCXlWf-css{background-color:#121212;border:1px solid #5E5E5E !important;border-radius:8px;color:#f2f2f2;font-family:var(--fonts-jetbrainsMono);font-size:var(--fontSizes-ms);height:unset;line-height:var(--lineHeights-body);padding:var(--space-md) var(--space-lg)}.c-dhzjXW-ieGaVeY-css{flex-grow:1}.c-dhzjXW-igEIieZ-css{height:100%;background:#121212;border:1px solid #5E5E5E !important;border-radius:8px}.c-dhzjXW-igEIieZ-css:hover{border-color:#8f8f8f !important}.c-dhzjXW-igEIieZ-css{color:#f2f2f2;font-size:var(--fontSizes-ms);padding:0;overflow:auto}.c-jgxVxd-ilgQfFT-css{grid-template-columns:repeat(2, minmax(0px, 1fr))}@media only screen and (max-width: 1300px){.c-jgxVxd-ilgQfFT-css{grid-template-columns:repeat(1, minmax(0px, 1fr))}}.c-jgxVxd-ihbthGk-css{width:100%;height:100%;max-height:80%;background-color:#292929;border:1px solid #5E5E5E !important;border-radius:8px;grid-template-columns:9fr 3fr}@media screen and (max-width: 1500px){.c-jgxVxd-ihbthGk-css{max-height:70%}}@media screen and (max-width: 1300px){.c-jgxVxd-ihbthGk-css{max-height:40%}}@media screen and (max-width: 851px){.c-jgxVxd-ihbthGk-css{max-height:100%;border:none !important;background-color:inherit;row-gap:var(--space-sm);grid-template-columns:repeat(1, minmax(0px, 1fr))}}.c-jgxVxd-idrAaWl-css{height:100%;overflow:hidden;border-top-left-radius:8px;border-bottom-left-radius:8px}@media screen and (max-width: 851px){.c-jgxVxd-idrAaWl-css{border-radius:8px}}.c-dhzjXW-ikdyFAG-css{background-color:#292929;flex:0 0 170px;padding:var(--space-md);border-end-end-radius:8px;border-start-end-radius:8px}@media screen and (max-width: 851px){.c-dhzjXW-ikdyFAG-css{border-radius:8px}}.c-dhzjXW-iculkRU-css{border-bottom:1px solid #5E5E5E !important}.c-hakyQ-idchnVT-css{background:#F2AA3C;height:var(--space-sm);width:var(--space-sm)}.c-kAArxJ-ifEHJXU-css{color:#E0E0E0}.c-hakyQ-iiHKkUW-css{background:#EC5E2A;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-ikGTPTu-css{background:#E2AA3A;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-ijVDGcn-css{background:#EA3A68;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-ibrsHGX-css{background:#E679D5;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-icInFh-css{background:#3F8F27;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-ihugmgO-css{background:#6CE9AE;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-idJRZry-css{background:#CCDC4B;height:var(--space-sm);width:var(--space-sm)}.c-hakyQ-ikYDmCv-css{background:#4DA9EE;height:var(--space-sm);width:var(--space-sm)}}</style></head><body class="bg-manitoulinLightBlack" data-scroll-lock="false"><main><template data-dgst="BAILOUT_TO_CLIENT_SIDE_RENDERING"></template></main><script>(self.__next_s=self.__next_s||[]).push(["https://cdn.cookielaw.org/scripttemplates/otSDKStub.js",{"data-document-language":"true","data-domain-script":"3e2b62ff-7ae7-4ac5-87c8-d5949ecafff5","type":"text/javascript"}])</script><script src="/_next/static/chunks/webpack-5cf9b03c4685f4fa.js" async=""></script><script>(self.__next_f=self.__next_f||[]).push([0]);self.__next_f.push([2,null])</script><script>self.__next_f.push([1,"1:HL[\"/_next/static/media/0064b95ffa9453c6-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n2:HL[\"/_next/static/media/46edd4ccaf5e549e-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n3:HL[\"/_next/static/media/4bac9051a9746ec4-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n4:HL[\"/_next/static/media/8b83042c50f11cab-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n5:HL[\"/_next/static/media/bb3ef058b751a6ad-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n6:HL[\"/_next/static/media/cc0cfc6c360d4093-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n7:HL[\"/_next/static/media/d102257ac65edb2a-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n8:HL[\"/_next/static/media/d4aa046a6a806d8f-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\n9:HL[\"/_next/static/media/db503b9c104834ef-s.p.woff2\",\"font\",{\"crossOrigin\":\"\",\"type\":\"font/woff2\"}]\na:HL[\"/_next/static/css/8f94c4dae2af4daf.css\",\"style\"]\nb:HL[\"/_next/static/css/9a3720f649c4b700.css\",\"style\"]\nc:HL[\"/_next/static/css/11122b5d7d585ae7.css\",\"style\"]\nd:HL[\"/_next/static/css/fd88663f7ac2a1fd.css\",\"style\"]\ne:HL[\"/_next/static/css/caf16cecf0d5eb2d.css\",\"style\"]\nf:HL[\"/_next/static/css/688be17216d781f9.css\",\"style\"]\n10:HL[\"/_next/static/css/b716460f5ba1aa64.css\",\"style\"]\n11:HL[\"/_next/static/css/15335b524c9ed37d.css\",\"style\"]\n12:HL[\"/_next/static/css/9ab03ffa4c735f8f.css\",\"style\"]\n"])</script><script>self.__next_f.push([1,"13:I[6849,[],\"\"]\n15:I[14593,[],\"\"]\n16:I[67159,[],\"\"]\n1d:I[52850,[],\"\"]\n1e:[]\n"])</script><script>self.__next_f.push([1,"0:[\"$\",\"$L13\",null,{\"buildId\":\"6ruTx8ZiNwzA8VA2zGzFt\",\"assetPrefix\":\"\",\"urlParts\":[\"\",\"?_rsc=1hiu2\"],\"initialTree\":[\"\",{\"children\":[\"(app-bar-layout)\",{\"children\":[\"explore\",{\"top\":[\"children\",{\"children\":[\"discover\",{\"children\":[\"__PAGE__\",{}]}]}],\"children\":[\"discover\",{\"children\":[\"__PAGE__\",{}]}]}]}]},\"$undefined\",\"$undefined\",true],\"initialSeedData\":[\"\",{\"children\":[\"(app-bar-layout)\",{\"children\":[\"explore\",{\"top\":[\"children\",{\"children\":[\"discover\",{\"children\":[\"__PAGE__\",{},[[\"$undefined\",\"$L14\",[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/9ab03ffa4c735f8f.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}]]],null],null]},[null,[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\",\"explore\",\"top\",\"children\",\"children\",\"discover\",\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"notFoundStyles\":\"$undefined\"}]],null]},[null,[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\",\"explore\",\"top\",\"children\",\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"notFoundStyles\":\"$undefined\"}]],null],\"children\":[\"discover\",{\"children\":[\"__PAGE__\",{},[[\"$L17\",\"$L18\",[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/9ab03ffa4c735f8f.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}]]],null],null]},[null,[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\",\"explore\",\"children\",\"discover\",\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"notFoundStyles\":\"$undefined\"}]],null]},[[[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/15335b524c9ed37d.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}]],\"$L19\"],null],null]},[[[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/11122b5d7d585ae7.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}],[\"$\",\"link\",\"1\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/fd88663f7ac2a1fd.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}],[\"$\",\"link\",\"2\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/caf16cecf0d5eb2d.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}],[\"$\",\"link\",\"3\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/688be17216d781f9.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}],[\"$\",\"link\",\"4\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/b716460f5ba1aa64.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}]],\"$L1a\"],null],null]},[[[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/8f94c4dae2af4daf.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}],[\"$\",\"link\",\"1\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/css/9a3720f649c4b700.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\"}]],\"$L1b\"],null],null],\"couldBeIntercepted\":false,\"initialHead\":[null,\"$L1c\"],\"globalErrorComponent\":\"$1d\",\"missingSlots\":\"$W1e\"}]\n"])</script><script>self.__next_f.push([1,"20:I[4790,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"12544\",\"static/chunks/12544-fac2301ea4b12d45.js\",\"30841\",\"static/chunks/30841-02381d56be476691.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"46128\",\"static/chunks/46128-e96fd11a1604d3a4.js\",\"39065\",\"static/chunks/39065-bd1cbbcb5fbe806d.js\",\"37373\",\"static/chunks/app/(app-bar-layout)/layout-e6b833b293a6a48d.js\"],\"AuthContextProvider\"]\n21:I[89518,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"stat"])</script><script>self.__next_f.push([1,"ic/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"303\",\"static/chunks/app/(app-bar-layout)/explore/layout-bfa292509e5202a5.js\"],\"default\"]\n22:I[25133,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"7095"])</script><script>self.__next_f.push([1,"3\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"12544\",\"static/chunks/12544-fac2301ea4b12d45.js\",\"30841\",\"static/chunks/30841-02381d56be476691.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"46128\",\"static/chunks/46128-e96fd11a1604d3a4.js\",\"39065\",\"static/chunks/39065-bd1cbbcb5fbe806d.js\",\"37373\",\"static/chunks/app/(app-bar-layout)/layout-e6b833b293a6a48d.js\"],\"VerificationBanner\"]\n23:I[85538,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7."])</script><script>self.__next_f.push([1,"js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"12544\",\"static/chunks/12544-fac2301ea4b12d45.js\",\"30841\",\"static/chunks/30841-02381d56be476691.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"46128\",\"static/chunks/46128-e96fd11a1604d3a4.js\",\"39065\",\"static/chunks/39065-bd1cbbcb5fbe806d.js\",\"37373\",\"static/chunks/app/(app-bar-layout)/layout-e6b833b293a6a48d.js\"],\"AppBar\"]\n25:I[65312,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"12544\",\"static/chunks/12544-fac2301ea4b12d45.js\",\"30841\",\"static/chunks/30841-02381d56be476691.j"])</script><script>self.__next_f.push([1,"s\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"46128\",\"static/chunks/46128-e96fd11a1604d3a4.js\",\"39065\",\"static/chunks/39065-bd1cbbcb5fbe806d.js\",\"37373\",\"static/chunks/app/(app-bar-layout)/layout-e6b833b293a6a48d.js\"],\"default\"]\n26:I[39133,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n27:I[69066,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f"])</script><script>self.__next_f.push([1,"42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"bodyClass\"]\n28:I[6892,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n29:\"$Sreact.suspense\"\n2a:I[34539,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703"])</script><script>self.__next_f.push([1,"ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n2b:I[46179,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js"])</script><script>self.__next_f.push([1,"\"],\"ThemeProvider\"]\n2c:I[97111,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"StarfleetProvider\"]\n2d:I[8777,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js"])</script><script>self.__next_f.push([1,"\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"SelectedModelProvider\"]\n2e:I[59524,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n2f:I[89404,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\""])</script><script>self.__next_f.push([1,"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"AdobeAnalytics\"]\n30:I[40733,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"\"]\n18:[[\"$\",\"h2\",null,{\"className\":\"mb-8 text-xl font-medium tracking-less\",\"children\":\"Discover\"}],\"$L1f\"]\n24:{\"killswitches/quick-actions.enabled\":true,\"killswitches/gpus-nav.enabled\":true,\"killswitches/remove-credits.enabled\":true,\"killswitches/org-selector.enabled\":true,\"killswitches/api-keys.enable"])</script><script>self.__next_f.push([1,"d\":true,\"killswitches/agents-nav.enabled\":false,\"killswitches/otp-verification.enabled\":true}\n"])</script><script>self.__next_f.push([1,"1a:[\"$\",\"$L20\",null,{\"children\":[[\"$\",\"$L21\",null,{\"flag\":\"killswitches/otp-verification.enabled\",\"initialFlags\":{\"killswitches/quick-actions.enabled\":true,\"killswitches/gpus-nav.enabled\":true,\"killswitches/remove-credits.enabled\":true,\"killswitches/org-selector.enabled\":true,\"killswitches/api-keys.enabled\":true,\"killswitches/agents-nav.enabled\":false,\"killswitches/otp-verification.enabled\":true},\"children\":[\"$\",\"$L22\",null,{}]}],[\"$\",\"$L23\",null,{\"initialFlags\":\"$24\"}],[\"$\",\"$L25\",null,{}],[\"$\",\"div\",null,{\"className\":\"flex flex-col items-center bg-manitoulinLightBlack px-4 pt-4 md:px-8\",\"children\":[[\"$\",\"div\",null,{\"className\":\"w-full\",\"children\":[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":[\"$\",\"div\",null,{\"className\":\"flex min-h-[75vh] flex-col items-center justify-center gap-md text-manitoulinLightWhite\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-lg font-medium\",\"children\":\"404 | Not Found\"}],[\"$\",\"$L26\",null,{\"href\":\"/\",\"children\":[\"Return Home\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row btn-tertiary btn-md btn-pill\"}]]}],\"notFoundStyles\":[]}]}],[\"$\",\"footer\",null,{\"className\":\"text-center\",\"children\":[[\"$\",\"div\",null,{\"className\":\"mx-auto mb-2.5 mt-[30px] flex flex-row flex-wrap justify-center gap-x-[20px] text-ms font-normal text-n300\",\"children\":[[\"$\",\"a\",null,{\"className\":\"hover:text-n000\",\"href\":\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA_Technology_Access_TOU.pdf\",\"children\":\"Terms of Use\"}],[\"$\",\"p\",null,{\"children\":\" | \"}],[\"$\",\"a\",null,{\"className\":\"hover:text-n000\",\"href\":\"https://www.nvidia.com/en-us/about-nvidia/privacy-policy/\",\"children\":\"Privacy Policy\"}],[\"$\",\"p\",null,{\"children\":\" | \"}],[\"$\",\"a\",null,{\"className\":\"hover:text-n000\",\"href\":\"https://www.nvidia.com/en-us/about-nvidia/privacy-center/\",\"children\":\"Manage My Privacy\"}],[\"$\",\"p\",null,{\"children\":\" | \"}],[\"$\",\"a\",null,{\"className\":\"hover:text-n000\",\"href\":\"https://developer.nvidia.com/contact\",\"children\":\"Contact\"}]]}],[\"$\",\"p\",null,{\"className\":\"mb-[30px] text-xs font-light text-manitoulinLightGray\",\"children\":[\"Copyright © \",2025,\" NVIDIA Corporation\"]}]]}]]}]]}]\n"])</script><script>self.__next_f.push([1,"1b:[\"$\",\"html\",null,{\"className\":\"__variable_d06c34 __variable_b20d9a kui-dark-theme bg-transparent font-sans\",\"id\":\"app\",\"lang\":\"en\",\"children\":[\"$\",\"body\",null,{\"className\":\"$27\",\"data-scroll-lock\":false,\"children\":[[\"$\",\"main\",null,{\"children\":[\"$\",\"$L28\",null,{\"children\":[\"$\",\"$29\",null,{\"children\":[\"$\",\"$L2a\",null,{\"children\":[\"$\",\"$L2b\",null,{\"children\":[\"$\",\"$L2c\",null,{\"children\":[\"$\",\"$L2d\",null,{\"children\":[\"$\",\"$L2e\",null,{\"children\":[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":[\"$\",\"div\",null,{\"className\":\"flex min-h-[75vh] flex-col items-center justify-center gap-md text-manitoulinLightWhite\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-lg font-medium\",\"children\":\"404 | Not Found\"}],[\"$\",\"$L26\",null,{\"href\":\"/\",\"children\":[\"Return Home\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row btn-tertiary btn-md btn-pill\"}]]}],\"notFoundStyles\":[]}]}]}]}]}]}]}]}]}],[\"$\",\"$L2f\",null,{}],[[\"$\",\"$L30\",null,{\"data-document-language\":\"true\",\"data-domain-script\":\"3e2b62ff-7ae7-4ac5-87c8-d5949ecafff5\",\"src\":\"https://cdn.cookielaw.org/scripttemplates/otSDKStub.js\",\"strategy\":\"beforeInteractive\",\"type\":\"text/javascript\"}],[\"$\",\"$L30\",null,{\"async\":true,\"id\":\"OptanonWrapper\",\"children\":\"function OptanonWrapper() {\\n const event = new Event('bannerLoaded');\\n window.dispatchEvent(event);\\n }\"}],[\"$\",\"$L30\",null,{\"async\":true,\"id\":\"GlobalPrivacyControl\",\"src\":\"https://images.nvidia.com/aem-dam/Solutions/ot-js/ot-custom.js\"}]]]}]}]\n1c:[[\"$\",\"meta\",\"0\",{\"name\":\"viewport\",\"content\":\"width=device-width, initial-scale=1\"}],[\"$\",\"meta\",\"1\",{\"charSet\":\"utf-8\"}],[\"$\",\"title\",\"2\",{\"children\":\"Try NVIDIA NIM APIs\"}],[\"$\",\"meta\",\"3\",{\"name\":\"description\",\"content\":\"Experience the leading models to build enterp"])</script><script>self.__next_f.push([1,"rise generative AI apps now.\"}],[\"$\",\"link\",\"4\",{\"rel\":\"canonical\",\"href\":\"https://build.nvidia.com\"}],[\"$\",\"meta\",\"5\",{\"property\":\"og:title\",\"content\":\"Try NVIDIA NIM APIs\"}],[\"$\",\"meta\",\"6\",{\"property\":\"og:description\",\"content\":\"Experience the leading models to build enterprise generative AI apps now.\"}],[\"$\",\"meta\",\"7\",{\"property\":\"og:image:type\",\"content\":\"image/jpeg\"}],[\"$\",\"meta\",\"8\",{\"property\":\"og:image:width\",\"content\":\"1200\"}],[\"$\",\"meta\",\"9\",{\"property\":\"og:image:height\",\"content\":\"630\"}],[\"$\",\"meta\",\"10\",{\"property\":\"og:image\",\"content\":\"https://build.nvidia.com/opengraph-image.jpg?6ec102a0470b935b\"}],[\"$\",\"meta\",\"11\",{\"property\":\"og:type\",\"content\":\"website\"}],[\"$\",\"meta\",\"12\",{\"name\":\"twitter:card\",\"content\":\"summary_large_image\"}],[\"$\",\"meta\",\"13\",{\"name\":\"twitter:title\",\"content\":\"Try NVIDIA NIM APIs\"}],[\"$\",\"meta\",\"14\",{\"name\":\"twitter:description\",\"content\":\"Experience the leading models to build enterprise generative AI apps now.\"}],[\"$\",\"meta\",\"15\",{\"name\":\"twitter:image:type\",\"content\":\"image/jpeg\"}],[\"$\",\"meta\",\"16\",{\"name\":\"twitter:image:width\",\"content\":\"1200\"}],[\"$\",\"meta\",\"17\",{\"name\":\"twitter:image:height\",\"content\":\"630\"}],[\"$\",\"meta\",\"18\",{\"name\":\"twitter:image\",\"content\":\"https://build.nvidia.com/twitter-image.jpg?6ec102a0470b935b\"}],[\"$\",\"link\",\"19\",{\"rel\":\"icon\",\"href\":\"/favicon.ico\",\"type\":\"image/x-icon\",\"sizes\":\"48x48\"}],[\"$\",\"meta\",\"20\",{\"name\":\"next-size-adjust\"}]]\n17:null\n"])</script><script>self.__next_f.push([1,"14:[[\"$\",\"div\",null,{\"className\":\"c-dhzjXW c-dhzjXW-irEjuD-align-stretch c-dhzjXW-iTKOFX-direction-column c-dhzjXW-awKDG-justify-flex-start c-dhzjXW-kVNAnR-wrap-nowrap flex flex-col gap-8 mx-auto mt-8 max-w-[1600px]\",\"data-testid\":\"kui-stack\",\"children\":[\"$L31\",[\"$\",\"div\",null,{\"aria-orientation\":\"horizontal\",\"className\":\"bg-n700 h-px w-full\",\"role\":\"separator\"}],\"$L32\"]}],\"$undefined\"]\n1f:[[\"$\",\"div\",null,{\"className\":\"c-dhzjXW c-dhzjXW-irEjuD-align-stretch c-dhzjXW-iTKOFX-direction-column c-dhzjXW-awKDG-justify-flex-start c-dhzjXW-kVNAnR-wrap-nowrap flex flex-col gap-8\",\"data-testid\":\"kui-stack\",\"children\":[\"$L33\",[\"$\",\"div\",null,{\"aria-orientation\":\"horizontal\",\"className\":\"bg-n700 h-px w-full\",\"role\":\"separator\"}],\"$L34\",[\"$\",\"div\",null,{\"aria-orientation\":\"horizontal\",\"className\":\"bg-n700 h-px w-full\",\"role\":\"separator\"}],\"$L35\",[\"$\",\"div\",null,{\"aria-orientation\":\"horizontal\",\"className\":\"bg-n700 h-px w-full\",\"role\":\"separator\"}],\"$L36\"]}],\"$undefined\"]\n"])</script><script>self.__next_f.push([1,"31:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L37\"}]}]]}]\n32:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L38\"}]}]]}]\n33:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L39\"}]}]]}]\n34:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L3a\"}]}]]}]\n35:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L3b\"}]}]]}]\n36:[\"$\",\"section\",null,{\"data-testid\":\"explore-carousel-view\",\"children\":[\"$undefined\",[\"$\",\"div\",null,{\"className\":\"relative\",\"children\":[\"$\",\"$29\",null,{\"children\":\"$L3c\"}]}]]}]\n"])</script><script>self.__next_f.push([1,"3d:I[19902,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/14195-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n51:I[32605,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"66697\",\"static/chunks/66697-63d4ca0711f91414.js\",\"19697\",\"static/chunks/19697-280b3bc2e9516441.js\",\"14195\",\"static/chunks/1419"])</script><script>self.__next_f.push([1,"5-243ccf9b70449304.js\",\"84327\",\"static/chunks/84327-0b3e44af5992e15b.js\",\"71931\",\"static/chunks/71931-dfd1072f67f502b5.js\",\"86916\",\"static/chunks/86916-670dd431128fa3c7.js\",\"63738\",\"static/chunks/app/(app-bar-layout)/explore/discover/page-f2a98f385e494f24.js\"],\"default\"]\n3e:T13da,"])</script><script>self.__next_f.push([1,"## Use Case Description\n\n[Physical AI](https://www.nvidia.com/en-us/glossary/generative-physical-ai/) is poised to transform manufacturing, supply chain and logistics—bringing unprecedented levels of industrial automation, intelligence, and autonomy to the world’s factories, warehouses, and industrial facilities.\n\nIn the smart factories and warehouses of today and of the future, humans and fleets of robots, including AGVs/AMRs, [humanoid robots](https://www.nvidia.com/en-us/use-cases/humanoid-robots/), intelligent cameras, and [visual AI agents](https://www.nvidia.com/en-us/autonomous-machines/intelligent-video-analytics-platform/) work together to achieve their objectives. To ensure their efficient operation in the real world, enterprises will rely on [digital twins](https://www.nvidia.com/en-us/glossary/digital-twin/) of their facilities to simulate interactions and performance of these different robot types and their objectives, ensuring they can work together seamlessly to accomplish their tasks.\n\nThe Mega NVIDIA Omniverse Blueprint, powered by [NVIDIA Omniverse™](https://www.nvidia.com/en-us/omniverse/), [OpenUSD](https://www.nvidia.com/en-us/omniverse/usd/), and [Isaac™ ROS](https://developer.nvidia.com/isaac/ros), enables enterprises to combine real-time [sensor simulation](https://www.nvidia.com/en-us/glossary/sensor-simulation/) and [synthetic data generation](https://www.nvidia.com/en-us/use-cases/synthetic-data/) to simulate these complex human-robot interactions and verify the performance of physical AI systems in [industrial digital twins](https://www.nvidia.com/en-us/use-cases/ai-for-virtual-factory-solutions/?deeplink=content-tab--1) before real-world deployment.\n\n## Experience Walkthrough\n\nWhen starting the experience, users are presented with a sample warehouse populated with racks, boxes, and autonomous mobile robots (AMRs) equipped with 3D LiDAR and RGB camera sensors.\n\nTo set up the simulation, users can select a location inside the warehouse and configure two AMRs. For each AMR configuration, users can:\n\n- Select either a “smart” robot that can detect and avoid obstacles on its path or a “simple” robot that can only follow preprogrammed paths. \n- Select from one of the four camera views (front, left, right, back) and the type of render they want to generate. \n- Create an AMR path to navigate interactively. \n\nOnce the AMRs are configured, the user clicks the “Run Simulation” button, and the AMR brain, World Simulator, and Sensor RTX™ service are deployed. As shown in the architectural diagram:\n\n- The AMR brains control the AMRs and send control signals to actuate them in the World Simulator. \n- The World Simulator runs the physics-based simulation of the AMR’s movement. \n- Each AMR has multiple sensors that are simulated: one 3D LiDAR, one IMU, and one RGB camera with multiple [AOVs](https://www.nvidia.com/en-us/on-demand/session/omniverse2020-om1458/), in addition to a top-view camera using NVIDIA Sensor RTX APIs. \n- Sensor data is streamed back to the AMR controllers to perceive the surroundings and determine the next step of control signals. \n\nNote that simulations typically take 15–20 minutes to complete. During periods of high demand, results may take longer to generate. Users receive a simulation ID (valid for 14 days) that allows them to return to the experience to view the results. While waiting for the simulation to complete, users receive periodic simulation progress updates and an [introduction video to the reference architecture](https://assets.ngc.nvidia.com/products/api-catalog/mega/explanation.mp4). To get more in-depth information about the blueprint, [read the technical blog](https://developer.nvidia.com/blog/simulating-robots-in-industrial-facility-digital-twins/).\n\n## Architecture Diagram\n![](https://assets.ngc.nvidia.com/products/api-catalog/mega/diagram.jpg)\n\n## What’s Included in the Blueprint\n\nNVIDIA Blueprints are comprehensive reference workflows designed to streamline AI application development across industries and accelerate deployment to production. Built with NVIDIA AI and Omniverse libraries, SDKs, and microservices, they provide a foundation for custom AI solutions. Each blueprint includes a reference code for constructing workflows, tools and documentation for deployment and customization, and a reference architecture outlining API definitions and microservice interoperability. By enabling rapid prototyping and speeding time to deployment, these blueprints empower enterprises to operationalize AI-driven solutions like AI agents, digital twins, synthetic data generation, and more.\n\n## Terms of Use\nGOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); use of the model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/)."])</script><script>self.__next_f.push([1,"3f:T125d,"])</script><script>self.__next_f.push([1,"### Use-Case Description\n\nImitation learning lets robots learn skills from observing human demonstrations. But gathering enough high-quality real-world datasets can be challenging, costly, and time-consuming. Synthetic data, generated from physically accurate simulations, addresses the challenge of limited real-world data acquisition by accelerating data collection and providing the diversity needed to generalize robot learning models.\n\nThe NVIDIA Isaac GR00T blueprint for synthetic manipulation motion generation is the ideal place to start. This is a reference workflow for creating exponentially large amounts of synthetic motion trajectories for robot manipulation from a small number of human demonstrations, built on NVIDIA Omniverse™ and NVIDIA Cosmos™.\n\nFirst, developers use a spatial computing device such as the Apple Vision Pro to portal into their simulated robot digital twin and record motion demonstration teleoperating a simulated robot. These recordings are then used to generate a larger set of physically accurate synthetic motion trajectories. Finally, the blueprint further augments the dataset by generating an exponentially large, photorealistic, and diverse set of training data. \n\n\u003cdiv style=\"background-color: #202020; padding: 16px; border-radius: 16px; color: #fff; line-height: 21px;\"\u003e\nNote: The first release of this blueprint is for single-arm manipulation only. Support for bi-manual humanoid robot manipulation is coming soon.\n\u003c/div\u003e\n\n### Experience Walkthrough\n\nThe overall experience is divided into four distinct parts: \n\n1\\. Choose from a pre-recorded set of human demonstrations.\n\n2\\. View the synthetically generated motion.\n\n3\\. Select from the list of pre-populated prompts to augment the generated motions.\n\n4\\. Click \"View Source Code” to retrieve the blueprint from GitHub.\n\n## Architecture Diagram\n![](https://assets.ngc.nvidia.com/products/api-catalog/gr00t/diagram.jpg)\n\n## What’s Included in the Blueprint\n\n##### Sample Recorded Data\n\n* Pre-recorded human demonstrations for a single-arm manipulation\n\n##### Robot Simulation and Training Frameworks\n\n* NVIDIA Isaac™ Lab, an open-source, unified framework for [robot learning](https://www.nvidia.com/en-us/glossary/robot-learning/) designed to help train robot policies built on Isaac Sim\n\n##### Data Generation\n\n* GR00T-Mimic, a feature in Isaac Lab, uses the recorded demonstrations as input to generate synthetic motion trajectories \n\n##### Data Augmentation\n\n* GR00T-Gen, a feature in Isaac Lab for augmenting 3D datasets to achieve the necessary photorealism and diversity \n* [Cosmos-Transfer1-7B](https://huggingface.co/nvidia/Cosmos-Transfer1-7B) model\n\n### File Deliverables\n\n**Input:** \n\n- Pre-selected collection of human demonstration recordings, captured with teleoperation in simulation \n- Pre-populated prompts to augment data with prepopulated prompts\n\n**Output:** \n\n- Synthetically-generated trajectories \n- Augmented video displayed on the screen\n- Jupyter notebook to recreate the end-to-end development experience\n \n#### Minimum System Requirements\nHardware Requirements\n\nGPU\n* NVIDIA 6000 Ada, 4090, 5090, L40, L40S, L20 and A40 or any higher level NVIDIA RTX™-capable GPU \n* Cosmos - HGX node (1x H100) TBD\n\nCPU\n* Intel Core i7 (7th Generation) \n* AMD Ryzen 5\n\n#### OS Requirements\n* Ubuntu 22.04 OS \n* Windows 11\n\n## Ethical Considerations\nNVIDIA believes trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting team to ensure the technologies meet requirements for the relevant industry and use case and address unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Licenses\n \nLicensing information for Isaac Lab can be found [here](https://isaac-sim.github.io/IsaacLab/main/source/refs/license.html).\n\nLicense information for For NVIDIA Cosmos: can be found under [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\n\nLicensing for GR00T-Mimic can be found [here](https://www.google.com/url?q=https://github.com/NVIDIA-Omniverse-blueprints/synthetic-manipulation-motion-generation\u0026sa=D\u0026source=docs\u0026ust=1741721959187379\u0026usg=AOvVaw2ACeuFe6HOCLbIpMpcnrKH)\n\n\n## Terms of Use\nGoverning Terms: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf)."])</script><script>self.__next_f.push([1,"40:T24d3,"])</script><script>self.__next_f.push([1,"# **Cosmos-Predict1**: A Suite of Diffusion-based World Foundation Models\n\n[**Cosmos**](https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6) | [**Code**](https://github.com/NVIDIA/Cosmos) | [**Paper**](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai)\n\n\n# Model Overview\n\n## Description:\n**Cosmos World Foundation Models**: A family of highly performant pre-trained world foundation models purpose-built for generating physics-aware videos and world states for physical AI development.\n\nThe Cosmos diffusion models are a collection of diffusion based world foundation models that generate dynamic, high quality videos from text, image, or video inputs. It can serve as the building block for various applications or research that are related to world generation. The models are ready for commercial use under NVIDIA Open Model license agreement.\n\n**Model Developer**: NVIDIA\n\n## Model Versions\n\nIn Cosmos 1.0 release, the Cosmos Diffusion WFM family includes the following models:\n- [Cosmos-Predict1-7B-Text2World](https://huggingface.co/nvidia/Cosmos-Predict1-7B-Text2World)\n - Given a text description, predict an output video of 121 frames.\n- [Cosmos-Predict1-14B-Text2World](https://huggingface.co/nvidia/Cosmos-Predict1-14B-Text2World)\n - Given a text description, predict an output video of 121 frames.\n- [Cosmos-Predict1-7B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-7B-Video2World)\n - Given a text description and an image as the first frame, predict the future 120 frames.\n- [Cosmos-Predict1-14B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-14B-Video2World)\n - Given a text description and an image as the first frame, predict the future 120 frames.\n\n\n### License:\nThis model is released under the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). For a custom license, please contact [cosmos-license@nvidia.com](mailto:cosmos-license@nvidia.com).\n\nUnder the NVIDIA Open Model License, NVIDIA confirms:\n\n* Models are commercially usable.\n* You are free to create and distribute Derivative Models.\n* NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.\n\n**Important Note**: If you bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or\nassociated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained\nin the Model, your rights under [NVIDIA Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/) will automatically terminate.\n\n## Model Architecture:\n\n**Cosmos-Predict1-7B-Text2World** and **Cosmos-Predict1-7B-Video2World** are diffusion transformer models designed for video denoising in the latent space. The network is composed of interleaved self-attention, cross-attention and feedforward layers as its building blocks. The cross-attention layers allow the model to condition on input text throughout the denoising process. Before each layers, adaptive layer normalization is applied to embed the time information for denoising. When image or video is provided as input, their latent frames are concatenated with the generated frames along the temporal dimension. Augment noise is added to conditional latent frames to bridge the training and inference gap.\n\n## Cosmos-Predict1-7B-Text2World Input/Output Specifications\n\n* **Input**\n\n * **Input Type(s)**: Text\n * **Input Format(s)**: String\n * **Input Parameters**: One-dimensional (1D)\n * **Other Properties Related to Input**:\n * The input string should contain fewer than 300 words and should provide descriptive content for world generation, such as a scene description, key objects or characters, background, and any specific actions or motions to be depicted within the 5-second duration.\n\n* **Output**\n * **Output Type(s)**: Video\n * **Output Format(s)**: mp4\n * **Output Parameters**: Three-dimensional (3D)\n * **Other Properties Related to Output**: The generated video will be a 5-second clip with a resolution of 1280x704 pixels at 24 frames per second (fps). The content of the video will visualize the input text description as a short animated scene, capturing the main elements mentioned in the input within the time constraints.\n\n## Cosmos-Predict1-7B-Video2World Input/Output Specifications\n\n* **Input**\n\n * **Input Type(s)**: Text+Image, Text+Video\n * **Input Format(s)**:\n * Text: String\n * Image: jpg, png, jpeg, webp\n * Video: mp4\n * **Input Parameters**:\n * Text: One-dimensional (1D)\n * Image: Two-dimensional (2D)\n * Video: Three-dimensional (3D)\n * **Other Properties Related to Input**:\n * The input string should contain fewer than 300 words and should provide descriptive content for world generation, such as a scene description, key objects or characters, background, and any specific actions or motions to be depicted within the 5-second duration.\n * The input image should be of 1280x704 resolution.\n * The input video should be of 1280x704 resolution and 9 input frames.\n\n* **Output**\n * **Output Type(s)**: Video\n * **Output Format(s)**: mp4\n * **Output Parameters**: Three-dimensional (3D)\n * **Other Properties Related to Output**: The generated video will be a 5-second clip with a resolution of 1280x704 pixels at 24 frames per second (fps). The content of the video will use the provided image as the first frame and visualize the input text description as a short animated scene, capturing the main elements mentioned in the input within the time constraints.\n\n## Software Integration\n**Runtime Engine(s):**\n* [Cosmos](https://github.com/NVIDIA/Cosmos)\n\n**Supported Hardware Microarchitecture Compatibility:**\n* NVIDIA Blackwell\n* NVIDIA Hopper\n* NVIDIA Ampere\n\n**Note**: We have only tested inference with BF16 precision.\n\n\n\n**Operating System(s):**\n* Linux (We have not tested on other operating systems.)\n\n\n# Usage\n\n* See [Cosmos](https://github.com/NVIDIA/Cosmos) for details.\n\n\n# Evaluation\n\nPlease see our [technical paper](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai) for detailed evaluations.\n\n## Inference Time and GPU Memory Usage\n\nThe numbers provided below may vary depending on system specs and are for reference only.\n\nWe report the maximum observed GPU memory usage during end-to-end inference. Additionally, we offer a series of model offloading strategies to help users manage GPU memory usage effectively.\n\nFor GPUs with limited memory (e.g., RTX 3090/4090 with 24 GB memory), we recommend fully offloading all models. For higher-end GPUs, users can select the most suitable offloading strategy considering the numbers provided below.\n\n### Cosmos-Predict1-7B-Text2World\n\n| Offloading Strategy | 7B Text2World | 14B Text2World |\n|-------------|---------|---------|\n| Offload prompt upsampler | 74.0 GB | \u003e 80.0 GB |\n| Offload prompt upsampler \u0026 guardrails | 57.1 GB | 70.5 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder | 38.5 GB | 51.9 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer | 38.3 GB | 51.7 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer \u0026 diffusion model | 24.4 GB | 39.0 GB |\n\nThe table below presents the end-to-end inference runtime on a single H100 GPU, excluding model initialization time.\n\n| 7B Text2World (offload prompt upsampler) | 14B Text2World (offload prompt upsampler, guardrails) |\n|---------|---------|\n| ~380 seconds | ~590 seconds |\n\n### Cosmos-Predict1-7B-Video2World\n\n| Offloading Strategy | 7B Video2World | 14B Video2World |\n|----------------------------------------------------------------------------------|---------|---------|\n| Offload prompt upsampler | 76.5 GB | \u003e 80.0 GB |\n| Offload prompt upsampler \u0026 guardrails | 59.9 GB | 73.3 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder | 41.3 GB | 54.8 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer | 41.1 GB | 54.5 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer \u0026 diffusion model | 27.3 GB | 39.0 GB |\n\nThe following table shows the end-to-end inference runtime on a single H100 GPU, excluding model initialization time:\n\n| 7B Video2World (offload prompt upsampler) | 14B Video2World (offload prompt upsampler, guardrails) |\n|---------|---------|\n| ~383 seconds | ~593 seconds |\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nFor more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety \u0026 Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"41:T6cb,Field | Response\n:------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\nIntended Application \u0026 Domain: | World Generation\nModel Type: | Transformer\nIntended Users: | Physical AI developers\nOutput: | Videos\nDescribe how the model works: | Generates videos based on video inputs\nTechnical Limitations: | The model may not follow the video input accurately.\nVerified to have met prescribed NVIDIA quality standards: | Yes\nPerformance Metrics: | Quantitative and Qualitative Evaluation\nPotential Known Risks: | The model's output can generate all forms of videos, including what may be considered toxic, offensive, or indecent.\nLicensing: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)42:T7ad,Field | Response\n:----------------------------------------------------------------------------------------------------------------------------------|:-----------------"])</script><script>self.__next_f.push([1,"------------------------------\nGeneratable or reverse engineerable personal information? | None Known\nProtected class data used to create this model? | None Known\nWas consent obtained for any personal data used? | None Known\nHow often is dataset reviewed? | Before Release\nIs a mechanism in place to honor data subject right of access or deletion of personal data? | Not Applicable\nIf personal data was collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable\nIf personal data was collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable\nIf personal data was collected for the development of this AI model, was it minimized to only what was required? | Not Applicable\nIs there provenance for all datasets used in training? | Yes\nDoes data labeling (annotation, metadata) comply with privacy laws? | Yes\nIs data compliant with data subject requests for data correction or removal, if such a request was made? | Not Applicable43:T719,invoke_url='https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b'\nfetch_url_format='https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/'\n\nauthorization_header='Authorization: Bearer $NVIDIA_API_KEY'\naccept_header='Accept: application/json'\ncontent_type_header='Content-Type: application/json'\n\ndata='{\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\":"])</script><script>self.__next_f.push([1," [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on the industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n}'\n\nresponse=$(curl --silent -i -w \"\\n%{http_code}\" --request POST \\\n --url \"$invoke_url\" \\\n --header \"$authorization_header\" \\\n --header \"$accept_header\" \\\n --header \"$content_type_header\" \\\n --data \"$data\"\n)\n\nhttp_code=$(echo \"$response\" | tail -n 1)\nreq_id=$(echo \"$response\" | grep -i '^nvcf-reqid:' | awk '{print $2}' | tr -d '\\r')\n\nwhile [ \"$http_code\" -eq 202 ]; do\n response=$(curl --silent -i -w \"\\n%{http_code}\" --request GET \\\n --url \"$fetch_url_format$req_id\" \\\n --header \"$authorization_header\" \\\n --header \"$accept_header\" \\\n --header \"$content_type_header\" \\\n )\n\n http_code=$(echo \"$response\" | tail -n 1)\n req_id=$(echo \"$response\" | grep -i '^nvcf-reqid:' | awk '{print $2}' | tr -d '\\r')\ndone\n\nif [ \"$http_code\" -ne 302 ]; then\n echo \"invocation failed with status $http_code\" \u003e\u00262\n echo \"$response\" \u003e\u00262\n exit 1\nfi\n\ndownload_url=$(echo \"$response\" | grep -i '^location:' | awk '{print $2}' | tr -d '\\r')\ncurl -L --output result.zip \"$download_url\"\n44:T5b2,import fs from \"fs\";\n\nconst invokeUrl = \"https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b\";\nconst fetchUrlFormat = \"https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/\";\n\nconst headers = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"application/json\",\n};\n\nconst payload = {\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\": [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on th"])</script><script>self.__next_f.push([1,"e industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n};\n\nlet response = await fetch(invokeUrl, {\n method: \"post\",\n body: JSON.stringify(payload),\n headers: { \"Content-Type\": \"application/json\", ...headers }\n});\n\nwhile (response.status === 202) {\n const requestId = response.headers.get(\"NVCF-REQID\");\n const fetchUrl = fetchUrlFormat + requestId;\n response = await fetch(fetchUrl, {\n method: \"get\",\n headers: headers\n });\n}\n\nif (response.status !== 200) {\n const errBody = await (await response.blob()).text();\n throw \"invocation failed with status \" + response.status + \" \" + errBody;\n}\n\nfs.writeFileSync('result.zip', Buffer.from(await response.arrayBuffer()));\n45:T4f1,import requests\n\ninvoke_url = \"https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b\"\nfetch_url_format = \"https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/\"\n\nheaders = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"application/json\",\n}\n\npayload = {\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\": [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on the industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n}\n\n# re-use connections\nsession = requests.Session()\n\nresponse = session.post(invoke_url, headers=headers, json=payload)\n\nwhile response.status_code == 202:\n request_id = response.headers.get(\"NVCF-REQID\")\n fetch_url = fetch_url_format + request_id\n response = session.get(fetch_url, headers=headers)\nresponse = requests.post(invoke_url, headers=headers, json=payload)\n\nr"])</script><script>self.__next_f.push([1,"esponse.raise_for_status()\n\nwith open('result.zip', 'wb') as f:\n f.write(response.content)\n46:T3958,"])</script><script>self.__next_f.push([1,"# NVIDIA Cosmos\n\nCosmos is NVIDIA’s World Foundation Model Development Platform that provides the tools to either finetune existing models or train new models from scratch.\n\n## Cosmos Model Family\n\nCosmos World Foundation Models (WFM) are a family of highly-performant pre-trained models purpose-built for generating physics-aware videos used for training robots. With Cosmos, developers can simulate a world in which robots function and train them to act and react responsibly in the real world before actual deployment.\n\nCosmos WFMs currently contain four main types of models: NeMo Curator, Cosmos Tokenizer, Cosmos Guardrail, and Cosmos World Foundation Model. NeMo Curator is a video curation pipeline that takes raw video frames, splits them into meaningful segments, and annotates them with semantic tags, object labels, and scene descriptions. The annotated images are then fed into the Cosmos Tokenizer, which produces a sequence of tokens. This step reduces data dimensionality enabling Cosmos World Foundation Model to effectively handle large or complex inputs for training. Cosmos WFM then consumes the curated/annotated video segments and learns the underlying physics and visual dynamics from real world data. When queried, Cosmos WFM outputs new token sequences that are then decoded back into high-resolution and physically realistic synthetic videos. Cosmos WFMs are pretrained on large-scale video datasets to expose them to a broad range of visual experiences, enabling them to serve as generalists. To construct a specialized WFM developers are expected to fine-tune Cosmos WFM using additional data collected from a specific use case. This additional data will help adapt Cosmos WFM to this intended use case, ensuring it can perform optimally under real-world conditions.\n\n## Specific Risk Areas and Mitigations\n\nWFMs can produce unrealistic outputs, generate unsafe content or may inadvertently amplify societal biases reflected in their training data. Collectively, these risks underscore the need for technical measures to mitigate risk and careful evaluation before leveraging Cosmos WFM in real-world applications.\n\n### **Cosmos Guardrail**\n\n\nFor the safe use of our world foundation models, we develop a comprehensive guardrail system. Cosmos Guardrail consists of two stages: the pre-Guard and the post-Guard stage. The pre-Guard stage leverages [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0), which is a fine-tuned version of [Llama-Guard](https://ai.meta.com/research/publications/llama-guard-llm-based-input-output-safeguard-for-human-ai-conversations/) trained on [NVIDIA’s Aegis Content Safety Dataset](https://huggingface.co/datasets/nvidia/Aegis-AI-Content-Safety-Dataset-1.0) and a blocklist filter that performs a lemmatized and whole-word keyword search to block harmful prompts. It then further sanitizes the user prompt by processing it through the Cosmos Text2World Prompt Upsampler. The post-Guard stage blocks harmful visual outputs using a video content safety classifier and a face blur filter.\n![](https://assets.ngc.nvidia.com/products/api-catalog/cosmos/system-card-diagram.jpg)\n\n\nCosmos pre-Guard first uses a simple blocklist-based checker for unsafe keyword detection. This is designed to block explicitly harmful generations by doing a keyword search on the prompt against a hard-coded blocklist of a large corpus of explicit and objectionable words. Input words are lemmatized using [WordNetLemmatizer](https://www.nltk.org/api/nltk.stem.WordNetLemmatizer.html?highlight=wordnet), a tool that uses a lexical database of the English language to extract the root word from its variants. These lemmatized words are then compared to the words in the hard-coded blocklist, and the entire prompt is rejected if any profanity is found.\n\nAs the second line of defense, Cosmos pre-Guard uses [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0) to detect unsafe content in semantically-complex prompts. Aegis is able to classify prompts into13 critical safety risk categories: violence, sexual, criminal planning, weapons, substance abuse, suicide, child sexual abuse material, hatred, harassment, threat, and profanity. If the input prompt is categorized as unsafe by this prompt filter, the video is not generated, and an error message is displayed. Any prompt that does not fall into the above categories is considered safe from the prompt-filtering standpoint.\n\nPrior to passing the prompt to the world generation models, the prompt is further augmented and indirectly sanitized via the Cosmos Text2World Prompt Upsampler. This is a bespoke model that not only compensates for the lack of specificity in the prompt but also steers clear of objectionable denotations or connotations.\n\nCosmos post-Guard is a vision-domain guardrail that is activated after the world content has been generated and comprises a video content safety filter and a face blur filter. Our video content safety filter used in the post-Guard stage has been trained on carefully-curated datasets and evaluated on human\\- annotated datasets created by Cosmos Red Team. To calibrate model outputs for the intended use case in the robotics and autonomous vehicle domains, we also automatically detect and blur all faces. We use RetinaFace, a state-of-the-art face detection model, to identify facial regions with high confidence scores. For any generated face region larger than 20 × 20 pixels, we apply pixelation to obscure features while preserving the overall scene composition. Note that by blurring all generated human faces in the video, potential biases based on age, gender, race and ethnicity in the output video are reduced.\n\n### **Balanced Datasets**\n\n\nCosmos WFM is trained using both proprietary and publicly available video datasets. We curated about 100M clips of videos ranging from 2 to 60 seconds from a 20M hour-long video collection. For each clip, we use a VLM ([13B-parameter VILA model](https://huggingface.co/Efficient-Large-Model/VILA-13b)) to provide a video caption per 256 frames. As our goal is to create a VLM that is able to generate physically realistic videos, we use the video captions to curate the training dataset to cover various physical applications:\n\n* Driving (11%),\n* Hand motion and object manipulation (16%),\n* Human motion and activity (10%),\n* Spatial awareness and navigation (16%),\n* First person point-of-view (8%),\n* Nature dynamics (20%),\n* Dynamic camera movements (8%),\n* Synthetically rendered (4%)\n* Others (7%)\n\nTo ensure effective distribution of the dataset we employ a taxonomy-based classifier to label video types and prune those that introduce unrealistic behaviors, such as purely animated or abstract patterns. Certain categories relevant to world foundation models (like human actions and interactions) are upsampled, while less critical ones (such as landscapes) are downsampled.\n\nA significant amount of the initial video data is either semantically redundant or contains different visual effects, which may induce unwanted artifacts in the generated videos if not appropriately handled. We therefore designed a sequence of data processing steps to find the most valuable parts of the raw videos for training. Shot boundary detection identifies where one shot ends and another begins, after which all footage is re-encoded into a uniform, high-quality MP4 format to ensure consistent loading and reduce codec discrepancies. The resulting video segments undergo several filtering processes. Motion filtering removes clips that are static or excessively shaky, and tags the remaining clips with camera motion types to enhance training signals. Visual quality filtering uses a video assessment model trained on [DOVER](https://github.com/VQAssessment/DOVER) to discard the bottom 15% in perceptual quality and applies an image aesthetic model exclude footage that is aesthetically poor. A deduplication step uses [InternVideo2](https://arxiv.org/abs/2403.15377) embeddings to identify near-duplicate content and preserves the highest-resolution version for minimal quality loss.\n\n## **Evaluation Methods**\n\nWe employ a dedicated red team to actively probe the system using both standard and adversarial examples that are collected in an internal attack prompt dataset. These video outputs are annotated by a team of expert annotators, who were specially trained for our task, to classify the generated video on a scale of 1-5 on multiple categories of harm related to the safety taxonomy. These annotations also specify the start and end-frames where the unsafe content is detected, thereby generating high-quality annotations. The red team also probed each guardrail component independently with targeted examples to identify weaknesses and improve performance in edge cases. As of the date of publication, the red team has tested and annotated over10, 000 distinct prompt-video pairs that were carefully crafted to cover a broad range of unsafe content. We separate out our safety testing into 4 categories:\n\n**Targeted unsafe testing**\n\nTargeted unsafe testing involves generating a corpus of manually curated unsafe prompts. These are intended to emulate common unsafe interactions that are performed by non-technical users of the system with basic or limited knowledge of multimodal AI attack vectors. These have a high likelihood of being caught by prompt filters, e.g. “Video of a naked person”.\n\n**Adversarial Attack Testing**\n\nAdversarial attack testing involves generating a corpus of unsafe prompts following the styles of AI attack published in literature. This type of testing will also leverage some (not all) of the prompts from content safety datasets like Aegis and automation tooling like [Garak](https://github.com/NVIDIA/garak).\n\n**Prompt Upsampler Toxicity**\n\nPrompt Upsampler Toxicity refers to a phenomenon where automated methods are used to “upsample” or expand a prompt by adding detail or context and inadvertently introduce unsafe content. Monitoring and mitigating Prompt Upsampler Toxicity ensures that content moderation systems remain effective throughout the entire generation pipeline, preserving user trust and upholding ethical standards.\n\n**Accidental Mishap Testing**\n\nAccidental mishap testing involves emulating the experience of a user prompting the model with a benign prompt, and getting unsafe content in return. This is the hardest category to test, since it does not have a fixed method or protocol for generation.\n\n## Governing Terms/Terms of Use\n\nAll Cosmos WFMs are deployed globally and are covered under NVIDIA’s [Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). This license agreement confirms that:\n\n* Models are commercially usable.\n* You are free to create and distribute derivative models.\n* NVIDIA does not claim ownership of any outputs generated using the models or derivative models.\n\nIf users bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained in the Model, the user’s rights under the NVIDIA Open Model License Agreement will automatically terminate. If users are interested in a custom license, they may contact cosmos-license@nvidia.com.\n\n## Deployment\n\nCosmos WFM is released under an open, permissive NVIDIA license, allowing users to download the model weights and run it on their own hardware. This means that developers can integrate the WFM into their existing workflows without dependency on external APIs. They can also tailor the model to specific domain needs, retrain or fine-tune the model with their private data. This approach fosters innovation especially for under-resourced stakeholders that cannot rely on paid services.\n\nOnce downloaded, NVIDIA has less visibility into how or where Cosmos is deployed, reducing opportunities to enforce content policies or guardrails. Downloadable models grant complete control to users, but also transfer responsibility to the users for preventing misuse, and implementing safety mechanisms, such as watermarking and content moderation. Watermarking in the context of WFMs is crucial to ensure traceability, and user awareness that generated content might not be authentic. Watermarks allow viewers and downstream users to identify AI-generated or AI-manipulated videos, helping prevent misinformation and misuse. Even though watermarking is typically the responsibility of the user, we still encourage the use of open-source libraries for watermarking by downstream users of Cosmos WFM. NVIDIA has actively promoted watermarking and has worked in consortiums and standards bodies to define common protocols for watermarking synthetic media.\n\nCosmos WFM is also hosted by NVIDIA at the NVIDIA API Catalog (build.nvidia.com) and accessible via a web-based user interface. In this case, NVIDIA manages infrastructure, updates, and safety features. End users with minimal machine learning expertise can harness powerful WFMs without worrying about infrastructure or setup. Hosted models give NVIDIA more oversight and moderation capabilities, for example:\n\n* Know Your Customer (KYC) and account verification ensures that users are who they claim to be, discouraging malicious actors and fostering accountability.\n* Usage monitoring securely records user activity and flags suspicious patterns, enabling traceability and compliance checks while helping identify harmful behavior.\n* Rate limiting prevents spamming and large-scale misuse, balancing computational resources and protecting against abuse or overwhelming the system.\n* Human review protocols provide an escalation path for questionable outputs or flagged user accounts. This is a dedicated moderation team for final decisions on content removals, user bans, or investigations.\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal supporting team to ensure this system meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\n## Getting Help/Support\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n"])</script><script>self.__next_f.push([1,"47:T2060,"])</script><script>self.__next_f.push([1,"# **Cosmos-Predict1**: A Suite of Autoregressive-based World Foundation Models\n\n[**Cosmos**](https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6) | [**Code**](https://github.com/NVIDIA/Cosmos) | [**Paper**](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai)\n\n\n# Model Overview\n\n## Description:\n**Cosmos World Foundation Models**: A family of highly performant pre-trained world foundation models purpose-built for generating physics-aware videos and world states for physical AI development.\n\nThe Cosmos autoregressive models are a collection of pre-trained world foundation models that are ideal for predicting and rapidly generating video sequences from video or image inputs for physical AI. They can serve as the building block for various applications or research that are related to world generation. The models are ready for commercial use under NVIDIA Open Model license agreement.\n\n**Model Developer**: NVIDIA\n\n## Model Versions\n\nIn Cosmos 1.0 release, the Cosmos Autoregressive WFM family includes the following models:\n- [Cosmos-Predict1-4B](https://huggingface.co/nvidia/Cosmos-Predict1-4B)\n - Given a 9-frame input video, predicts the future 24 frames.\n - Given an image as the first frame, predicts the future 32 frames.\n- [Cosmos-Predict1-5B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-5B-Video2World)\n - Given text description and a 9-frame input video, predicts the future 24 frames.\n - Given text description and an image as the first frame, predicts the future 32 frames.\n- [Cosmos-Predict1-12B](https://huggingface.co/nvidia/Cosmos-Predict1-12B)\n - Given a 9-frame input video, predicts the future 24 frames.\n - Given an image as the first frame, predicts the future 32 frames.\n- [Cosmos-Predict1-13B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-13B-Video2World)\n - Given text description and a 9-frame input video, predicts the future 24 frames.\n - Given text description and an image as the first frame, predicts the future 32 frames.\n\n### License:\nThis model is released under the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license). For a custom license, please contact [cosmos-license@nvidia.com](mailto:cosmos-license@nvidia.com).\n\nUnder the NVIDIA Open Model License, NVIDIA confirms:\n\n* Models are commercially usable.\n* You are free to create and distribute Derivative Models.\n* NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.\n\n**Important Note**: If you bypass, disable, reduce the efficacy of, or circumvent any technical limitation, **safety guardrail** or\nassociated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained\nin the Model, your rights under [NVIDIA Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license) will automatically terminate.\n* [Cosmos-1.0-Guardrail](https://huggingface.co/nvidia/Cosmos-1.0-Guardrail) is the safety guardrail for this model.\n\n## Model Architecture:\n\nCosmos-Predict1-5B-Video2World is an autoregressive transformer model designed for world generation. The network is composed of interleaved self-attention, cross-attention, and feedforward layers as its building blocks. The cross-attention layers allow the model to condition on input text throughout the decoding process.\n\n## Input/Output Specifications\n\n* **Input**\n\n * **Input Type(s)**: Text+Image, Text+Video\n * **Input Format(s)**:\n * Text: String\n * Image: jpg, png, jpeg, webp\n * Video: mp4\n * **Input Parameters**:\n * Text: One-dimensional (1D)\n * Image: Two-dimensional (2D)\n * Video: Three-dimensional (3D)\n * **Other Properties Related to Input**:\n * The input string should contain fewer than 300 words and should provide descriptive content for world generation, such as a scene description, key objects or characters, background, and any specific actions or motions to be depicted within the 1-second duration.\n * The input image and video should be of 1024x640 resolution.\n\n* **Output**\n * **Output Type(s)**: Video\n * **Output Format(s)**: mp4\n * **Output Parameters**: Three-dimensional (3D)\n * **Other Properties Related to Output**:\n * For text+image input, the generated video will be a 32-frame clip with a resolution of 1024x640 pixels, conditioned on the input image as the first video frame.\n * For text+video input, the generated video will be a 24-frame clip with a resolution of 1024x640 pixels, conditioned on the first 9 frames of the input video.\n * The content of the video will visualize the input text description as a short animated scene, capturing the main elements mentioned in the input.\n\n## Software Integration:\n**Runtime Engine(s):**\n* [Cosmos](https://github.com/NVIDIA/Cosmos)\n\n**Supported Hardware Microarchitecture Compatibility:**\n* NVIDIA Blackwell\n* NVIDIA Hopper\n* NVIDIA Ampere\n\n**Note**: We have only tested doing inference with BF16 precision.\n\n\n\n**Operating System(s):**\n* Linux (We have not tested on other operating systems.)\n\n\n# Usage\n\n* See [Cosmos](https://github.com/NVIDIA/Cosmos) for details.\n\n\n# Evaluation\n\nPlease see our [technical paper](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai) for detailed evaluations.\n\n## Inference Time and GPU Memory Usage\n\nThese numbers may vary based on system specifications and are provided for reference only.\n\n| Offloading Strategy | Cosmos-Predict1-5B-Video2World | Cosmos-Predict1-13B-Video2World |\n|-------------|---------|---------|\n| No offloading | 66.2 GB | \u003e 80 GB |\n| Guardrails | 58.7 GB | 76.6 GB |\n| Guardrails \u0026 T5 encoder | 41.3 GB | 58.0 GB |\n| Guardrails \u0026 T5 encoder \u0026 Diffusion decoder | 29.0 GB | 46.9 GB |\n| Guardrails \u0026 T5 encoder \u0026 Diffusion decoder \u0026 Tokenizer | 28.8 GB | 46.7 GB |\n| Guardrails \u0026 T5 encoder \u0026 Diffusion decoder \u0026 Tokenizer \u0026 AR model | 21.1 GB | 30.9 GB |\n\nEnd-to-end inference runtime on one H100 with no offloading for 5B model and guardrail offloading for 13B, after model initialization:\n\n| Cosmos-Predict1-5B-Video2World | Cosmos-Predict1-13B-Video2World |\n|---------|---------|\n| ~73 seconds | ~150 seconds |\n\n## Failure Analysis\n\nOur models now support video extension up to 33 frames. Starting from either a single image or a 9-frame video input, it can generate the remaining frames to reach the 33-frame length (generating 32 or 24 frames respectively).\n\nWe have evaluated all eight possible configurations (4 models × 2 vision input types: image or video) using 100 test videos from physical AI domains. Below are the failure rates for each configuration:\n\n| Model | Image input | Video input (9 frames) |\n|:------------------------------------------|:--------------:|:-------------------------:|\n| Cosmos-Predict1-4B | 15% | 1% |\n| Cosmos-Predict1-5B-Video2World | 7% | 2% |\n| Cosmos-Predict1-12B | 2% | 1% |\n| Cosmos-Predict1-13B-Video2World | 3% | 0% |\n\nWe define failure cases as videos with severe distortions, such as:\n\n* Sudden appearance of large unexpected objects\n* Video degrading to a single solid color\n\nNote that the following are not considered failures in our analysis:\n\n* Static video frames\n* Minor object distortions or artifacts\n\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nFor more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety \u0026 Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"48:T6cb,Field | Response\n:------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\nIntended Application \u0026 Domain: | World Generation\nModel Type: | Transformer\nIntended Users: | Physical AI developers\nOutput: | Videos\nDescribe how the model works: | Generates videos based on video inputs\nTechnical Limitations: | The model may not follow the video input accurately.\nVerified to have met prescribed NVIDIA quality standards: | Yes\nPerformance Metrics: | Quantitative and Qualitative Evaluation\nPotential Known Risks: | The model's output can generate all forms of videos, including what may be considered toxic, offensive, or indecent.\nLicensing: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)49:T7ad,Field | Response\n:----------------------------------------------------------------------------------------------------------------------------------|:-----------------"])</script><script>self.__next_f.push([1,"------------------------------\nGeneratable or reverse engineerable personal information? | None Known\nProtected class data used to create this model? | None Known\nWas consent obtained for any personal data used? | None Known\nHow often is dataset reviewed? | Before Release\nIs a mechanism in place to honor data subject right of access or deletion of personal data? | Not Applicable\nIf personal data was collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable\nIf personal data was collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable\nIf personal data was collected for the development of this AI model, was it minimized to only what was required? | Not Applicable\nIs there provenance for all datasets used in training? | Yes\nDoes data labeling (annotation, metadata) comply with privacy laws? | Yes\nIs data compliant with data subject requests for data correction or removal, if such a request was made? | Not Applicable4a:T3958,"])</script><script>self.__next_f.push([1,"# NVIDIA Cosmos\n\nCosmos is NVIDIA’s World Foundation Model Development Platform that provides the tools to either finetune existing models or train new models from scratch.\n\n## Cosmos Model Family\n\nCosmos World Foundation Models (WFM) are a family of highly-performant pre-trained models purpose-built for generating physics-aware videos used for training robots. With Cosmos, developers can simulate a world in which robots function and train them to act and react responsibly in the real world before actual deployment.\n\nCosmos WFMs currently contain four main types of models: NeMo Curator, Cosmos Tokenizer, Cosmos Guardrail, and Cosmos World Foundation Model. NeMo Curator is a video curation pipeline that takes raw video frames, splits them into meaningful segments, and annotates them with semantic tags, object labels, and scene descriptions. The annotated images are then fed into the Cosmos Tokenizer, which produces a sequence of tokens. This step reduces data dimensionality enabling Cosmos World Foundation Model to effectively handle large or complex inputs for training. Cosmos WFM then consumes the curated/annotated video segments and learns the underlying physics and visual dynamics from real world data. When queried, Cosmos WFM outputs new token sequences that are then decoded back into high-resolution and physically realistic synthetic videos. Cosmos WFMs are pretrained on large-scale video datasets to expose them to a broad range of visual experiences, enabling them to serve as generalists. To construct a specialized WFM developers are expected to fine-tune Cosmos WFM using additional data collected from a specific use case. This additional data will help adapt Cosmos WFM to this intended use case, ensuring it can perform optimally under real-world conditions.\n\n## Specific Risk Areas and Mitigations\n\nWFMs can produce unrealistic outputs, generate unsafe content or may inadvertently amplify societal biases reflected in their training data. Collectively, these risks underscore the need for technical measures to mitigate risk and careful evaluation before leveraging Cosmos WFM in real-world applications.\n\n### **Cosmos Guardrail**\n\n\nFor the safe use of our world foundation models, we develop a comprehensive guardrail system. Cosmos Guardrail consists of two stages: the pre-Guard and the post-Guard stage. The pre-Guard stage leverages [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0), which is a fine-tuned version of [Llama-Guard](https://ai.meta.com/research/publications/llama-guard-llm-based-input-output-safeguard-for-human-ai-conversations/) trained on [NVIDIA’s Aegis Content Safety Dataset](https://huggingface.co/datasets/nvidia/Aegis-AI-Content-Safety-Dataset-1.0) and a blocklist filter that performs a lemmatized and whole-word keyword search to block harmful prompts. It then further sanitizes the user prompt by processing it through the Cosmos Text2World Prompt Upsampler. The post-Guard stage blocks harmful visual outputs using a video content safety classifier and a face blur filter.\n![](https://assets.ngc.nvidia.com/products/api-catalog/cosmos/system-card-diagram.jpg)\n\n\nCosmos pre-Guard first uses a simple blocklist-based checker for unsafe keyword detection. This is designed to block explicitly harmful generations by doing a keyword search on the prompt against a hard-coded blocklist of a large corpus of explicit and objectionable words. Input words are lemmatized using [WordNetLemmatizer](https://www.nltk.org/api/nltk.stem.WordNetLemmatizer.html?highlight=wordnet), a tool that uses a lexical database of the English language to extract the root word from its variants. These lemmatized words are then compared to the words in the hard-coded blocklist, and the entire prompt is rejected if any profanity is found.\n\nAs the second line of defense, Cosmos pre-Guard uses [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0) to detect unsafe content in semantically-complex prompts. Aegis is able to classify prompts into13 critical safety risk categories: violence, sexual, criminal planning, weapons, substance abuse, suicide, child sexual abuse material, hatred, harassment, threat, and profanity. If the input prompt is categorized as unsafe by this prompt filter, the video is not generated, and an error message is displayed. Any prompt that does not fall into the above categories is considered safe from the prompt-filtering standpoint.\n\nPrior to passing the prompt to the world generation models, the prompt is further augmented and indirectly sanitized via the Cosmos Text2World Prompt Upsampler. This is a bespoke model that not only compensates for the lack of specificity in the prompt but also steers clear of objectionable denotations or connotations.\n\nCosmos post-Guard is a vision-domain guardrail that is activated after the world content has been generated and comprises a video content safety filter and a face blur filter. Our video content safety filter used in the post-Guard stage has been trained on carefully-curated datasets and evaluated on human\\- annotated datasets created by Cosmos Red Team. To calibrate model outputs for the intended use case in the robotics and autonomous vehicle domains, we also automatically detect and blur all faces. We use RetinaFace, a state-of-the-art face detection model, to identify facial regions with high confidence scores. For any generated face region larger than 20 × 20 pixels, we apply pixelation to obscure features while preserving the overall scene composition. Note that by blurring all generated human faces in the video, potential biases based on age, gender, race and ethnicity in the output video are reduced.\n\n### **Balanced Datasets**\n\n\nCosmos WFM is trained using both proprietary and publicly available video datasets. We curated about 100M clips of videos ranging from 2 to 60 seconds from a 20M hour-long video collection. For each clip, we use a VLM ([13B-parameter VILA model](https://huggingface.co/Efficient-Large-Model/VILA-13b)) to provide a video caption per 256 frames. As our goal is to create a VLM that is able to generate physically realistic videos, we use the video captions to curate the training dataset to cover various physical applications:\n\n* Driving (11%),\n* Hand motion and object manipulation (16%),\n* Human motion and activity (10%),\n* Spatial awareness and navigation (16%),\n* First person point-of-view (8%),\n* Nature dynamics (20%),\n* Dynamic camera movements (8%),\n* Synthetically rendered (4%)\n* Others (7%)\n\nTo ensure effective distribution of the dataset we employ a taxonomy-based classifier to label video types and prune those that introduce unrealistic behaviors, such as purely animated or abstract patterns. Certain categories relevant to world foundation models (like human actions and interactions) are upsampled, while less critical ones (such as landscapes) are downsampled.\n\nA significant amount of the initial video data is either semantically redundant or contains different visual effects, which may induce unwanted artifacts in the generated videos if not appropriately handled. We therefore designed a sequence of data processing steps to find the most valuable parts of the raw videos for training. Shot boundary detection identifies where one shot ends and another begins, after which all footage is re-encoded into a uniform, high-quality MP4 format to ensure consistent loading and reduce codec discrepancies. The resulting video segments undergo several filtering processes. Motion filtering removes clips that are static or excessively shaky, and tags the remaining clips with camera motion types to enhance training signals. Visual quality filtering uses a video assessment model trained on [DOVER](https://github.com/VQAssessment/DOVER) to discard the bottom 15% in perceptual quality and applies an image aesthetic model exclude footage that is aesthetically poor. A deduplication step uses [InternVideo2](https://arxiv.org/abs/2403.15377) embeddings to identify near-duplicate content and preserves the highest-resolution version for minimal quality loss.\n\n## **Evaluation Methods**\n\nWe employ a dedicated red team to actively probe the system using both standard and adversarial examples that are collected in an internal attack prompt dataset. These video outputs are annotated by a team of expert annotators, who were specially trained for our task, to classify the generated video on a scale of 1-5 on multiple categories of harm related to the safety taxonomy. These annotations also specify the start and end-frames where the unsafe content is detected, thereby generating high-quality annotations. The red team also probed each guardrail component independently with targeted examples to identify weaknesses and improve performance in edge cases. As of the date of publication, the red team has tested and annotated over10, 000 distinct prompt-video pairs that were carefully crafted to cover a broad range of unsafe content. We separate out our safety testing into 4 categories:\n\n**Targeted unsafe testing**\n\nTargeted unsafe testing involves generating a corpus of manually curated unsafe prompts. These are intended to emulate common unsafe interactions that are performed by non-technical users of the system with basic or limited knowledge of multimodal AI attack vectors. These have a high likelihood of being caught by prompt filters, e.g. “Video of a naked person”.\n\n**Adversarial Attack Testing**\n\nAdversarial attack testing involves generating a corpus of unsafe prompts following the styles of AI attack published in literature. This type of testing will also leverage some (not all) of the prompts from content safety datasets like Aegis and automation tooling like [Garak](https://github.com/NVIDIA/garak).\n\n**Prompt Upsampler Toxicity**\n\nPrompt Upsampler Toxicity refers to a phenomenon where automated methods are used to “upsample” or expand a prompt by adding detail or context and inadvertently introduce unsafe content. Monitoring and mitigating Prompt Upsampler Toxicity ensures that content moderation systems remain effective throughout the entire generation pipeline, preserving user trust and upholding ethical standards.\n\n**Accidental Mishap Testing**\n\nAccidental mishap testing involves emulating the experience of a user prompting the model with a benign prompt, and getting unsafe content in return. This is the hardest category to test, since it does not have a fixed method or protocol for generation.\n\n## Governing Terms/Terms of Use\n\nAll Cosmos WFMs are deployed globally and are covered under NVIDIA’s [Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). This license agreement confirms that:\n\n* Models are commercially usable.\n* You are free to create and distribute derivative models.\n* NVIDIA does not claim ownership of any outputs generated using the models or derivative models.\n\nIf users bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained in the Model, the user’s rights under the NVIDIA Open Model License Agreement will automatically terminate. If users are interested in a custom license, they may contact cosmos-license@nvidia.com.\n\n## Deployment\n\nCosmos WFM is released under an open, permissive NVIDIA license, allowing users to download the model weights and run it on their own hardware. This means that developers can integrate the WFM into their existing workflows without dependency on external APIs. They can also tailor the model to specific domain needs, retrain or fine-tune the model with their private data. This approach fosters innovation especially for under-resourced stakeholders that cannot rely on paid services.\n\nOnce downloaded, NVIDIA has less visibility into how or where Cosmos is deployed, reducing opportunities to enforce content policies or guardrails. Downloadable models grant complete control to users, but also transfer responsibility to the users for preventing misuse, and implementing safety mechanisms, such as watermarking and content moderation. Watermarking in the context of WFMs is crucial to ensure traceability, and user awareness that generated content might not be authentic. Watermarks allow viewers and downstream users to identify AI-generated or AI-manipulated videos, helping prevent misinformation and misuse. Even though watermarking is typically the responsibility of the user, we still encourage the use of open-source libraries for watermarking by downstream users of Cosmos WFM. NVIDIA has actively promoted watermarking and has worked in consortiums and standards bodies to define common protocols for watermarking synthetic media.\n\nCosmos WFM is also hosted by NVIDIA at the NVIDIA API Catalog (build.nvidia.com) and accessible via a web-based user interface. In this case, NVIDIA manages infrastructure, updates, and safety features. End users with minimal machine learning expertise can harness powerful WFMs without worrying about infrastructure or setup. Hosted models give NVIDIA more oversight and moderation capabilities, for example:\n\n* Know Your Customer (KYC) and account verification ensures that users are who they claim to be, discouraging malicious actors and fostering accountability.\n* Usage monitoring securely records user activity and flags suspicious patterns, enabling traceability and compliance checks while helping identify harmful behavior.\n* Rate limiting prevents spamming and large-scale misuse, balancing computational resources and protecting against abuse or overwhelming the system.\n* Human review protocols provide an escalation path for questionable outputs or flagged user accounts. This is a dedicated moderation team for final decisions on content removals, user bans, or investigations.\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal supporting team to ensure this system meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\n## Getting Help/Support\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n"])</script><script>self.__next_f.push([1,"4b:T142f,"])</script><script>self.__next_f.push([1,"#### Experience a Real-time Wind Tunnel\n\nThis experience shows an interactive virtual wind tunnel. Simulating airflow in a virtual wind tunnel with computational fluid dynamics (CFD) requires millions of complex calculations. Without AI, it can take minutes to see the result of a single design change. Developers building the next generation of AI-powered CAE tools are combining simulation AI with immersive virtual environments to enable real-time digital twins where design changes instantly update in the simulation, as you see in this demonstration. NVIDIA is making it easier to create interactive design tools by introducing Omniverse Blueprint for interactive aerodynamics.\n\nTo enable real-time performance in a virtual wind tunnel, simulation AI models are first trained offline on representative datasets. For this demo, the training dataset was created using [Luminary Cloud’s GPU-accelerated CFD solver](https://www.luminarycloud.com/), which models complex airflow over diverse geometries. The simulation AI learns the complex relationships between car geometries (STLs) and airflow. This blueprint is compatible with industry-standard CFD solvers and can connect to third-party tools for meshing and geometry morphing, creating watertight meshes for simulation-ready geometries.\n\nThe blueprint also integrates NVIDIA PhysicsNeMo (our framework for simulation AI) with CFD solver data. This enables developers to train surrogate models from scratch or fine-tune NIM™ foundation models, reducing AI training time and cost. Once trained, the AI runs simulations orders of magnitude faster than traditional CFD, enabling real-time aerodynamic flow simulation. This speed provides designers with creative freedom, allowing designers to innovate and explore changes interactively.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/digital-twins-for-fluid-simulation/diagram.jpg)\n\n## What’s Included in the Blueprint\nNVIDIA Blueprints are comprehensive reference workflows designed to streamline AI application development across industries and accelerate deployment to production. Built with NVIDIA AI and Omniverse libraries, SDKs, and microservices, they provide a foundation for custom AI solutions. Each blueprint includes reference code for constructing workflows, tools, and documentation for deployment and customization, and a reference architecture outlining API definitions and microservice interoperability. By enabling rapid prototyping and speeding time to deployment, these blueprints empower enterprises to operationalize AI-driven solutions like AI agents, digital twins, and synthetic data generation, and more.\n\n### Included NIM Microservices\n* domino-automotive-aero 1.0\n\n### World State Controller\nThis reference solution implements an Omniverse Kit application controller that maintains the application world state and connects the 3D world stage with simulation results produced by the surrogate model.\n\nIn summary, this NVIDIA Omniverse Blueprint offers a starting point for building real-time digital twins for computer-aided engineering (CAE) workflows combining CUDA-X™ accelerated solvers, NVIDIA PhysicsNeMo for simulation AI, and Omniverse for high-quality rendering.\n\n## Minimum System Requirements\n\n#### Hardware Requirements\nThe real-time wind tunnel blueprint supports the following hardware:\n\n* At least 2x RTX™ GPUs with at least 40GB of memory each, e.g., 2xL40S or 2xA6000 \n* 128GB RAM \n* 32 CPU Cores \n* 100 GB Storage\n\n#### OS Requirements\n* Linux \\- Ubuntu 22.04 or 24.04\n\n#### Software Requirements\n* Git: For version control and repository management. \n* Git Large File System (LFS): For large files that are too large to efficiently store in a Git repository. \n* Python 3: For scripting and automation. \n* Docker: For containerized development and deployment. Ensure non-root users have Docker permissions. \n* NVIDIA Container Toolkit: For GPU-accelerated containerized development and deployment. Installation and configuring docker steps are required.\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Terms of Use\nGOVERNING TERMS: The software and materials are governed by the NVIDIA Software License Agreement (found at https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the Product-Specific Terms for NVIDIA Omniverse (found at NVIDIA Agreements | Enterprise Software | Product Specific Terms for Omniverse)."])</script><script>self.__next_f.push([1,"4c:{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"type\":\"blueprint\"}\n4d:{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"type\":\"blueprint\"}\n4e:{\"name\":\"cosmos-predict1-7b\",\"type\":\"model\"}\n4f:{\"name\":\"cosmos-predict1-5b\",\"type\":\"model\"}\n50:{\"name\":\"digital-twins-for-fluid-simulation\",\"type\":\"blueprint\"}\n"])</script><script>self.__next_f.push([1,"3a:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"displayName\":\"Test Multi-Robot Fleets for Industrial Automation\",\"publisher\":\"nvidia\",\"shortDescription\":\"Simulate, test, and optimize physical AI and robotic fleets at scale in industrial digital twins before real-world deployment.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/mega-multi-robot-fleets-for-industrial-automation.jpg\",\"labels\":[\"Blueprint\",\"NVIDIA Omniverse\",\"industrial\",\"omniverse blueprint\",\"simulation\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"true\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:09:17.934Z\",\"description\":\"$3e\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-31T16:01:59.380Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"efdb7052-704c-49eb-bd69-a854d79ecc21\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Example - AI Virtual Assistant for Customer Service\",\"description\":\"This API schema describes all the endpoints exposed by the AI Virtual Assistant for Customer Service NIM Blueprint\",\"version\":\"1.0.0\"},\"paths\":{\"/agent/metrics\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Get Metrics\",\"operationId\":\"get_metrics_agent_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/agent/health\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Health Check\",\"description\":\"Perform a Health Check\\n\\nReturns 200 when service is up. This does not check the health of downstream services.\",\"operationId\":\"health_check_agent_health_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthResponse\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}}}},\"/agent/generate\":{\"post\":{\"tags\":[\"Agent\"],\"summary\":\"Generate Response\",\"description\":\"Generate and stream the response to the provided prompt.\",\"operationId\":\"generate_response_agent_generate_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}},\"x-nvai-meta\":{\"examples\":[{\"name\":\"What was my last purchase?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What was my last purchase?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"},{\"name\":\"What product did I order first?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What product did I order first?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"}]}}}},\"components\":{\"schemas\":{\"AgentRequest\":{\"properties\":{\"messages\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":50000},{\"type\":\"null\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between user and assistant. The last input message should have role user. A message with the the system role is optional, and must be the very first message if it is present.\",\"default\":[]},\"user_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"User Id\",\"description\":\"A unique identifier representing your end-user.\",\"default\":\"\"},\"session_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\",\"default\":\"\"},\"api_type\":{\"type\":\"string\",\"title\":\"Api Type\",\"description\":\"The type of API action: 'create_session', 'end_session', 'generate', or 'summary'.\"}},\"type\":\"object\",\"required\":[\"api_type\"],\"title\":\"AgentRequest\",\"description\":\"Definition of the Prompt API data type.\"},\"AgentResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"maxLength\":100000,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Id\",\"default\":\"\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/AgentResponseChoices\"},\"type\":\"array\",\"maxItems\":256,\"title\":\"Choices\",\"default\":[]},\"session_id\":{\"type\":\"string\",\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\"}},\"type\":\"object\",\"title\":\"AgentResponse\",\"description\":\"Definition of Chain APIs resopnse data type\"},\"AgentResponseChoices\":{\"properties\":{\"index\":{\"type\":\"integer\",\"maximum\":256,\"minimum\":0,\"format\":\"int64\",\"title\":\"Index\",\"default\":0},\"message\":{\"$ref\":\"#/components/schemas/Message\",\"default\":{\"role\":\"assistant\",\"content\":\"\"}},\"finish_reason\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Finish Reason\",\"default\":\"\"},\"sentiment\":{\"type\":\"string\",\"title\":\"Sentiment\",\"description\":\"Any sentiment associated with this message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"AgentResponseChoices\",\"description\":\"Definition of Chain response choices\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthResponse\":{\"properties\":{\"message\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"HealthResponse\"},\"Message\":{\"properties\":{\"role\":{\"type\":\"string\",\"maxLength\":256,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Role\",\"description\":\"Role for a message AI, User and System\",\"default\":\"user\"},\"content\":{\"type\":\"string\",\"maxLength\":131072,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Content\",\"description\":\"The input query/prompt to the pipeline.\",\"default\":\"I am going to Paris, what should I see?\"}},\"type\":\"object\",\"title\":\"Message\",\"description\":\"Definition of the Chat Message type.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}},\"tags\":[{\"name\":\"Health\",\"description\":\"APIs for checking and monitoring server liveliness and readiness.\"},{\"name\":\"Agent\",\"description\":\"Core APIs for interacting with the agent.\"}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-31T16:01:59.959Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:09:18.304Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; use of the model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License Agreement\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Notify When Available\",\"url\":\"https://www.nvidia.com/en-us/omniverse/mega-multi-robot-fleets-notify-me/\"}},\"artifactName\":\"mega-multi-robot-fleets-for-industrial-automation\"},\"config\":{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"displayName\":\"Synthetic Manipulation Motion Generation for Robotics\",\"publisher\":\"nvidia\",\"shortDescription\":\"Generate exponentially large amounts of synthetic motion trajectories for robot manipulation from just a few human demonstrations.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/isaac-gr00t-synthetic-manipulation.jpg\",\"labels\":[\"Blueprint\",\"Humanoids\",\"NVIDIA Isaac GR00T\",\"image-to-world\",\"physical ai\",\"robot learning\",\"robotics\",\"synthetic data\",\"teleop\",\"text-to-world\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"true\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:11:29.060Z\",\"description\":\"$3f\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-19T18:02:34.533Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"93003d0b-a573-4906-99ef-82a0f14cf24d\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Example - AI Virtual Assistant for Customer Service\",\"description\":\"This API schema describes all the endpoints exposed by the AI Virtual Assistant for Customer Service NIM Blueprint\",\"version\":\"1.0.0\"},\"paths\":{\"/agent/metrics\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Get Metrics\",\"operationId\":\"get_metrics_agent_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/agent/health\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Health Check\",\"description\":\"Perform a Health Check\\n\\nReturns 200 when service is up. This does not check the health of downstream services.\",\"operationId\":\"health_check_agent_health_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthResponse\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}}}},\"/agent/generate\":{\"post\":{\"tags\":[\"Agent\"],\"summary\":\"Generate Response\",\"description\":\"Generate and stream the response to the provided prompt.\",\"operationId\":\"generate_response_agent_generate_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}},\"x-nvai-meta\":{\"examples\":[{\"name\":\"What was my last purchase?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What was my last purchase?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"},{\"name\":\"What product did I order first?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What product did I order first?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"}]}}}},\"components\":{\"schemas\":{\"AgentRequest\":{\"properties\":{\"messages\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":50000},{\"type\":\"null\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between user and assistant. The last input message should have role user. A message with the the system role is optional, and must be the very first message if it is present.\",\"default\":[]},\"user_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"User Id\",\"description\":\"A unique identifier representing your end-user.\",\"default\":\"\"},\"session_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\",\"default\":\"\"},\"api_type\":{\"type\":\"string\",\"title\":\"Api Type\",\"description\":\"The type of API action: 'create_session', 'end_session', 'generate', or 'summary'.\"}},\"type\":\"object\",\"required\":[\"api_type\"],\"title\":\"AgentRequest\",\"description\":\"Definition of the Prompt API data type.\"},\"AgentResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"maxLength\":100000,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Id\",\"default\":\"\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/AgentResponseChoices\"},\"type\":\"array\",\"maxItems\":256,\"title\":\"Choices\",\"default\":[]},\"session_id\":{\"type\":\"string\",\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\"}},\"type\":\"object\",\"title\":\"AgentResponse\",\"description\":\"Definition of Chain APIs resopnse data type\"},\"AgentResponseChoices\":{\"properties\":{\"index\":{\"type\":\"integer\",\"maximum\":256,\"minimum\":0,\"format\":\"int64\",\"title\":\"Index\",\"default\":0},\"message\":{\"$ref\":\"#/components/schemas/Message\",\"default\":{\"role\":\"assistant\",\"content\":\"\"}},\"finish_reason\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Finish Reason\",\"default\":\"\"},\"sentiment\":{\"type\":\"string\",\"title\":\"Sentiment\",\"description\":\"Any sentiment associated with this message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"AgentResponseChoices\",\"description\":\"Definition of Chain response choices\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthResponse\":{\"properties\":{\"message\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"HealthResponse\"},\"Message\":{\"properties\":{\"role\":{\"type\":\"string\",\"maxLength\":256,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Role\",\"description\":\"Role for a message AI, User and System\",\"default\":\"user\"},\"content\":{\"type\":\"string\",\"maxLength\":131072,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Content\",\"description\":\"The input query/prompt to the pipeline.\",\"default\":\"I am going to Paris, what should I see?\"}},\"type\":\"object\",\"title\":\"Message\",\"description\":\"Definition of the Chat Message type.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}},\"tags\":[{\"name\":\"Health\",\"description\":\"APIs for checking and monitoring server liveliness and readiness.\"},{\"name\":\"Agent\",\"description\":\"Core APIs for interacting with the agent.\"}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-19T18:02:35.234Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:11:29.392Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; use of the model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License Agreement\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-Omniverse-blueprints/synthetic-motion-generation\"}},\"artifactName\":\"isaac-gr00t-synthetic-manipulation\"},\"config\":{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"cosmos-predict1-7b\",\"displayName\":\"cosmos-predict1-7b\",\"publisher\":\"nvidia\",\"shortDescription\":\"Generates physics-aware video world states from text and image prompts for physical AI development.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/cosmos-predict1-7b.jpg\",\"labels\":[\"Physical AI\",\"image-to-world\",\"robotics\",\"text-to-world\",\"Synthetic Data Generation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"Field | Response\\n:---------------------------------------------------------------------------------------------------|:---------------\\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None\\nMeasures taken to mitigate against unwanted bias: | None\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:09:32.918Z\",\"description\":\"$40\",\"explainability\":\"$41\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$42\",\"safetyAndSecurity\":\"Field | Response\\n:---------------------------------------------------|:----------------------------------\\nModel Application(s): | World Generation\\nDescribe the life critical impact (if present). | None Known\\nUse Case Restrictions: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\\nModel and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog.\",\"updatedDate\":\"2025-03-18T19:58:04.974Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"b7daa55b-4865-402a-98ef-11f9f86a4873\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim for more details.\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"license\":{\"name\":\"TODO\",\"url\":\"https://todo\"},\"title\":\"NVIDIA NIM API for nvidia/cosmos-1.0-diffusion-7b-text2world\",\"version\":\"1.0.0\"},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"v1/infer\":{\"post\":{\"summary\":\" Infer\",\"operationId\":\"_infer_infer_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Text2WorldRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Text2WorldResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"templates\":[{\"requestEjs\":{\"curl\":\"$43\",\"node.js\":\"$44\",\"python\":\"$45\"}}]}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Text2WorldRequest\":{\"additionalProperties\":false,\"properties\":{\"prompt\":{\"description\":\"The prompt to use for generation. The prompt describes what should be present in the output generation.\",\"title\":\"Prompt\",\"type\":\"string\"},\"seed\":{\"anyOf\":[{\"exclusiveMaximum\":4294967296,\"minimum\":0,\"type\":\"integer\"},{\"type\":\"null\"}],\"default\":0,\"description\":\"The seed which governs generation. Changing the seed with other inputs fixed results in different outputs.\",\"title\":\"Seed\"}},\"required\":[\"prompt\"],\"title\":\"Text2WorldRequest\",\"type\":\"object\"},\"Text2WorldResponse\":{\"properties\":{\"asset_url\":{\"description\":\"The URL of the generated video (mp4)\",\"examples\":[\"https://api.nvcf.nvidia.com/v1/assets/12345\"],\"title\":\"Asset Url\",\"type\":\"string\"}},\"required\":[\"asset_url\"],\"title\":\"Text2WorldResponse\",\"type\":\"object\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}},\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}}},\"security\":[{\"Token\":[]}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:58:05.525Z\",\"nvcfFunctionId\":\"01327741-a1cb-4bdb-a31e-5391c8ca48c2\",\"createdDate\":\"2025-03-18T19:09:33.259Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-cosmos-1_0-diffusion-7b\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS:\u003c/b\u003e This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA API Trial Terms of Service\u003c/u\u003e\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA Open Model License Agreement\u003c/u\u003e\u003c/a\u003e (v. December 20, 2024).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Download and Post-Train\",\"url\":\"$undefined\"},\"fineTuneModal\":{\"ngc\":\"https://catalog.ngc.nvidia.com/orgs/nvidia/teams/cosmos/collections/cosmos\",\"huggingFace\":\"https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6\",\"nemo\":\"https://github.com/NVIDIA/Cosmos\"},\"maxRequests\":20,\"systemCard\":\"$46\"},\"artifactName\":\"cosmos-predict1-7b\"},\"config\":{\"name\":\"cosmos-predict1-7b\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"cosmos-predict1-5b\",\"displayName\":\"cosmos-predict1-5b\",\"publisher\":\"nvidia\",\"shortDescription\":\"Generates future frames of a physics-aware world state based on simply an image or short video prompt for physical AI development.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/cosmos-predict1-5b.jpg\",\"labels\":[\"Physical AI\",\"policy evaluation\",\"robotics\",\"Synthetic Data Generation\",\"video-to-world\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"Field | Response\\n:---------------------------------------------------------------------------------------------------|:---------------\\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None\\nMeasures taken to mitigate against unwanted bias: | None\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:09:32.317Z\",\"description\":\"$47\",\"explainability\":\"$48\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$49\",\"safetyAndSecurity\":\"Field | Response\\n:---------------------------------------------------|:----------------------------------\\nModel Application(s): | World Generation\\nDescribe the life critical impact (if present). | None Known\\nUse Case Restrictions: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\\nModel and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog.\",\"updatedDate\":\"2025-03-18T19:58:06.053Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"baf7b21c-af4d-4219-87fd-683ff0cd6ac2\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"FastAPI\",\"version\":\"0.1.0\",\"description\":\"FastAPI is a modern, fast (high-performance), web framework for building APIs with Python 3.6+ based on standard Python type hints.\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"Name\",\"url\":\"https://place\"}},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"/infer\":{\"post\":{\"summary\":\" Infer\",\"operationId\":\"_infer_infer_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Video2WorldRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Video2WorldResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}}}},\"/v2/health/ready\":{\"get\":{\"summary\":\"Health\",\"operationId\":\"health_v2_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"},\"Video2WorldRequest\":{\"additionalProperties\":false,\"properties\":{\"prompt\":{\"description\":\"The prompt to use for generation. The prompt describes what should be present in the output generation.\",\"title\":\"Prompt\",\"type\":\"string\"},\"input_image_index\":{\"anyOf\":[{\"maximum\":1,\"minimum\":0,\"type\":\"integer\"},{\"type\":\"null\"}],\"default\":0,\"description\":\"The index of the input image to use for generation. The input image is used as a reference for the generation.\",\"title\":\"Input Image Index\"},\"seed\":{\"anyOf\":[{\"exclusiveMaximum\":4294967296,\"minimum\":0,\"type\":\"integer\"},{\"type\":\"null\"}],\"default\":0,\"description\":\"The seed which governs generation. Changing the seed with other inputs fixed results in different outputs.\",\"title\":\"Seed\"}},\"required\":[\"prompt\"],\"title\":\"Video2WorldRequest\",\"type\":\"object\"},\"Video2WorldResponse\":{\"properties\":{\"asset_url\":{\"description\":\"The URL of the generated video (mp4)\",\"examples\":[\"https://api.nvcf.nvidia.com/v1/assets/12345\"],\"title\":\"Asset Url\",\"type\":\"string\"}},\"required\":[\"asset_url\"],\"title\":\"Video2WorldResponse\",\"type\":\"object\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:58:06.697Z\",\"nvcfFunctionId\":\"eef816a3-3940-413b-93c9-513ae29f34f9\",\"createdDate\":\"2025-03-18T19:09:32.632Z\",\"attributes\":{\"apiDocsUrl\":null,\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS:\u003c/b\u003e This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA API Trial Terms of Service\u003c/u\u003e\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA Open Model License Agreement\u003c/u\u003e\u003c/a\u003e (v. December 20, 2024).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Download and Post-Train\",\"url\":\"$undefined\"},\"fineTuneModal\":{\"ngc\":\"https://catalog.ngc.nvidia.com/orgs/nvidia/teams/cosmos/collections/cosmos\",\"huggingFace\":\"https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6\",\"nemo\":\"https://github.com/NVIDIA/Cosmos\"},\"maxRequests\":20,\"systemCard\":\"$4a\"},\"artifactName\":\"cosmos-predict1-5b\"},\"config\":{\"name\":\"cosmos-predict1-5b\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"digital-twins-for-fluid-simulation\",\"displayName\":\"Build a Digital Twin for Interactive Fluid Simulation\",\"publisher\":\"nvidia\",\"shortDescription\":\"This NVIDIA Omniverse™ Blueprint demonstrates how commercial software vendors can create interactive digital twins.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/digital-twins-for-fluid-simulation.jpg\",\"labels\":[\"Blueprint\",\"CAE\",\"Computer-aided-engineering\",\"External Aerodynamics\",\"NVIDIA Omniverse\",\"simulation\"],\"attributes\":[{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-11-18T18:51:39.577Z\",\"description\":\"$4b\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T20:09:15.623Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"31cd111e-f34f-4870-928b-b6f927bff889\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Build a Digital Twin for Interactive Fluid Simulation\",\"description\":\"TBD\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"Name\",\"url\":\"https://place\"}},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"/nvidia/digital-twins-for-fluid-simulation\":{\"post\":{\"summary\":\"Search Post\",\"operationId\":\"search_post_nvidia_usdcava_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/DeepSearchSearchRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Search results\",\"content\":{\"application/json\":{\"schema\":{\"items\":{\"$ref\":\"#/components/schemas/SearchResult\"},\"type\":\"array\",\"title\":\"Response Search Post V2 USD Search Post\"},\"example\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"402\":{\"description\":\"Payment Required\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/TBD\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}},\"x-nvai-meta\":{\"templates\":[{\"requestEjs\":{\"python\":\"import requests\\nimport json\\n\\nresponse = requests.post(\\n url='https://ai.api.nvidia.com/v1/nvidia/digital-twins-for-fluid-simulation', \\n headers={\\n 'Authorization': 'Bearer $NVIDIA_API_KEY',\\n 'Accept': 'application/json',\\n 'Content-Type': 'application/json',\\n },\\n data=json.dumps(\\n dict(\\n description = 'pallet', \\n file_extension_include ='usd*', \\n return_images ='true', \\n return_metadata = 'true', \\n return_vision_generated_metadata = 'true', \\n cutoff_threshold = '1.05', \\n limit = '50'\\n )\\n )\\n)\\n\\nprint(response.status_code)\\nprint(json.dumps(response.json(), indent=2))\",\"shell\":\"curl --request POST --url 'https://ai.api.nvidia.com/v1/nvidia/digital-twins-for-fluid-simulation' --header 'Authorization: Bearer $NVIDIA_API_KEY' --header 'Accept: application/json' --header 'Content-Type: application/json' --data '\\n{\\n 'description': 'pallet',\\n 'file_extension_include': 'usd*',\\n 'return_images': 'true',\\n 'return_metadata': 'true',\\n 'return_vision_generated_metadata': 'true',\\n 'cutoff_threshold': '1.05',\\n 'limit': '50'\\n}'\\n\"}}]}}},\"security\":[{\"Token\":[]}],\"components\":{\"schemas\":{\"DeepSearchSearchRequest\":{\"properties\":{\"description\":{\"type\":\"string\",\"title\":\"Description\",\"description\":\"Conduct text-based searches powered by AI\",\"default\":\"\"},\"image_similarity_search\":{\"type\":\"string\",\"title\":\"Image Similarity Search\",\"description\":\"Perform similarity searches based on images\",\"default\":\"\"},\"file_name\":{\"type\":\"string\",\"title\":\"File Name\",\"description\":\"Filter results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"exclude_file_name\":{\"type\":\"string\",\"title\":\"Exclude File Name\",\"description\":\"Exclude results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_include\":{\"type\":\"string\",\"title\":\"File Extension Include\",\"description\":\"Filter results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_exclude\":{\"type\":\"string\",\"title\":\"File Extension Exclude\",\"description\":\"Exclude results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"created_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created After\",\"description\":\"Filter results to only include assets created after a specified date\",\"default\":\"\"},\"created_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created Before\",\"description\":\"Filter results to only include assets created before a specified date\",\"default\":\"\"},\"modified_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified After\",\"description\":\"Filter results to only include assets modified after a specified date\",\"default\":\"\"},\"modified_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified Before\",\"description\":\"Filter results to only include assets modified before a specified date\",\"default\":\"\"},\"file_size_greater_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Greater Than\",\"description\":\"Filter results to only include files larger than a specific size\",\"default\":\"\"},\"file_size_less_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Less Than\",\"description\":\"Filter results to only include files smaller than a specific size\",\"default\":\"\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\",\"description\":\"Filter results to only include assets created by a specific user. In case AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_created_by\":{\"type\":\"string\",\"title\":\"Exclude Created By\",\"description\":\"Exclude assets created by a specific user from the results\",\"default\":\"\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\",\"description\":\"Filter results to only include assets modified by a specific user. In the case, when AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_modified_by\":{\"type\":\"string\",\"title\":\"Exclude Modified By\",\"description\":\"Exclude assets modified by a specific user from the results\",\"default\":\"\"},\"similarity_threshold\":{\"type\":\"number\",\"maximum\":2,\"minimum\":0,\"title\":\"Similarity Threshold\",\"description\":\"Set the similarity threshold for embedding-based searches. This functionality allows filterring duplicates and returning only those results that are different from each other. Assets are considered to be duplicates if the cosine distance betwen the embeddings a smaller than the similarity_threshold value, which could be in the [0, 2] range.\"},\"cutoff_threshold\":{\"type\":\"number\",\"minimum\":0,\"title\":\"Cutoff Threshold\",\"description\":\"Set the cutoff threshold for embedding-based searches\"},\"search_path\":{\"type\":\"string\",\"title\":\"Search Path\",\"description\":\"Specify the search path within the storage backend. This path should not contain the storage backend URL, just the asset path on the storage backend. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"search_in_scene\":{\"type\":\"string\",\"title\":\"Search In Scene\",\"description\":\"Conduct the search within a specific scene. Provide the full URL for the asset including the storage backend URL prefix.\",\"default\":\"\"},\"filter_by_properties\":{\"type\":\"string\",\"title\":\"Filter By Properties\",\"description\":\"Filter assets by USD attributes where at least one root prim matches (note: only supported for a subset of attributes indexed). Format: `attribute1=abc,attribute2=456`\",\"default\":\"\"},\"min_bbox_x\":{\"type\":\"number\",\"title\":\"Min Bbox X\",\"description\":\"Filter by minimum X axis dimension of the asset's bounding box\"},\"min_bbox_y\":{\"type\":\"number\",\"title\":\"Min Bbox Y\",\"description\":\"Filter by minimum Y axis dimension of the asset's bounding box\"},\"min_bbox_z\":{\"type\":\"number\",\"title\":\"Min Bbox Z\",\"description\":\"Filter by minimum Z axis dimension of the asset's bounding box\"},\"max_bbox_x\":{\"type\":\"number\",\"title\":\"Max Bbox X\",\"description\":\"Filter by maximum X axis dimension of the asset's bounding box\"},\"max_bbox_y\":{\"type\":\"number\",\"title\":\"Max Bbox Y\",\"description\":\"Filter by maximum Y axis dimension of the asset's bounding box\"},\"max_bbox_z\":{\"type\":\"number\",\"title\":\"Max Bbox Z\",\"description\":\"Filter by maximum Z axis dimension of the asset's bounding box\"},\"return_images\":{\"type\":\"boolean\",\"title\":\"Return Images\",\"description\":\"Return images if set to True\",\"default\":false},\"return_metadata\":{\"type\":\"boolean\",\"title\":\"Return Metadata\",\"description\":\"Return metadata if set to True\",\"default\":false},\"return_root_prims\":{\"type\":\"boolean\",\"title\":\"Return Root Prims\",\"description\":\"Return root prims if set to True\",\"default\":false},\"return_predictions\":{\"type\":\"boolean\",\"title\":\"Return Predictions\",\"description\":\"Return predictions if set to True\",\"default\":false},\"return_in_scene_instances_prims\":{\"type\":\"boolean\",\"title\":\"Return In Scene Instances Prims\",\"description\":\"[in-scene search only] Return prims of instances of objects found in the scene\",\"default\":false},\"embedding_knn_search_method\":{\"allOf\":[{\"$ref\":\"#/components/schemas/SearchMethod\"}],\"description\":\"Search method, approximate should be faster but is less accurate. Default is exact\"},\"limit\":{\"type\":\"integer\",\"maximum\":10000,\"minimum\":1,\"title\":\"Limit\",\"description\":\"Set the maximum number of results to return from the search, default is 32\"},\"vision_metadata\":{\"type\":\"string\",\"title\":\"Vision Metadata\",\"description\":\"Input string for the vision metadata query\"},\"return_vision_generated_metadata\":{\"type\":\"boolean\",\"title\":\"Return Vision Generated Metadata\",\"description\":\"Return vision generated metadata if True\",\"default\":false}},\"type\":\"object\",\"title\":\"DeepSearchSearchRequest\",\"example\":{\"description\":\"box\",\"return_metadata\":true,\"limit\":16}},\"EdgeType\":{\"type\":\"string\",\"enum\":[\"depends_on\",\"parent_prim\",\"root_prim\",\"source_asset\"],\"title\":\"EdgeType\",\"description\":\"An enumeration.\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"minimum\":100,\"maximum\":999,\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"format\":\"^.{1, 1024}$\",\"maxLength\":1024,\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"format\":\"^.{1, 256}$\",\"maxLength\":256,\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36,\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"PaymentRequiredError\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"},\"HealthResponse\":{\"properties\":{\"health_code\":{\"type\":\"string\",\"title\":\"Health Code\",\"description\":\"Status code, OK if successful, otherwise error code.\",\"examples\":[\"200\"]},\"health_code_description\":{\"type\":\"string\",\"title\":\"Health Code Description\",\"description\":\"Status description.\",\"examples\":[\"OK\"]},\"headers\":{\"type\":\"object\",\"title\":\"Headers\",\"description\":\"Response headers.\",\"default\":{\"cache-control\":[\"no-cache\",\"no-store\",\"max-age=0\",\"must-revalidate\"],\"expires\":0,\"pragma\":\"no-cache\"}}},\"type\":\"object\",\"required\":[\"health_code\",\"health_code_description\"],\"title\":\"HealthResponse\"},\"Metadata\":{\"properties\":{\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"}},\"type\":\"object\",\"title\":\"Metadata\"},\"Path\":{\"properties\":{\"version\":{\"type\":\"integer\",\"title\":\"Version\"},\"type\":{\"$ref\":\"#/components/schemas/PathType\"},\"uri\":{\"type\":\"string\",\"title\":\"Uri\"},\"acl\":{\"items\":{\"$ref\":\"#/components/schemas/PathPermission\"},\"type\":\"array\"},\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"},\"event\":{\"$ref\":\"#/components/schemas/PathEvent\"},\"mounted\":{\"type\":\"boolean\",\"title\":\"Mounted\"},\"transaction_id\":{\"type\":\"string\",\"title\":\"Transaction Id\"},\"destination\":{\"type\":\"string\",\"title\":\"Destination\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"is_deleted\":{\"type\":\"boolean\",\"title\":\"Is Deleted\"},\"deleted_by\":{\"type\":\"string\",\"title\":\"Deleted By\"},\"deleted_timestamp\":{\"type\":\"number\",\"title\":\"Deleted Timestamp\"}},\"type\":\"object\",\"title\":\"Path\"},\"PathEvent\":{\"enum\":[\"full\",\"create\",\"update\",\"delete\",\"change_acl\",\"set_path_options\",\"lock\",\"unlock\",\"rename\",\"copy\",\"replace_version\"],\"title\":\"PathEvent\",\"description\":\"An enumeration.\"},\"PathPermission\":{\"enum\":[\"read\",\"write\",\"admin\"],\"title\":\"PathPermission\",\"description\":\"An enumeration.\"},\"PathType\":{\"enum\":[\"\",\"asset\",\"folder\",\"channel\",\"mount\",\"object\",\"none\"],\"title\":\"PathType\",\"description\":\"An enumeration.\"},\"Prediction\":{\"properties\":{\"tag\":{\"type\":\"string\",\"title\":\"Tag\"},\"prob\":{\"type\":\"number\",\"title\":\"Prob\"}},\"type\":\"object\",\"required\":[\"tag\",\"prob\"],\"title\":\"Prediction\"},\"Prim\":{\"properties\":{\"scene_url\":{\"type\":\"string\",\"title\":\"Scene Url\"},\"usd_path\":{\"type\":\"string\",\"title\":\"Usd Path\"},\"prim_type\":{\"type\":\"string\",\"title\":\"Prim Type\"},\"bbox_max\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Max\"},\"bbox_min\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Min\"},\"bbox_midpoint\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Midpoint\"},\"bbox_dimension_x\":{\"type\":\"number\",\"title\":\"Bbox Dimension X\"},\"bbox_dimension_y\":{\"type\":\"number\",\"title\":\"Bbox Dimension Y\"},\"bbox_dimension_z\":{\"type\":\"number\",\"title\":\"Bbox Dimension Z\"},\"source_asset_url\":{\"type\":\"string\",\"title\":\"Source Asset Url\"},\"properties\":{\"type\":\"object\",\"title\":\"Properties\"}},\"type\":\"object\",\"required\":[\"scene_url\",\"usd_path\",\"prim_type\",\"bbox_max\",\"bbox_min\",\"bbox_midpoint\",\"bbox_dimension_x\",\"bbox_dimension_y\",\"bbox_dimension_z\"],\"title\":\"Prim\"},\"SearchItem\":{\"properties\":{\"status\":{\"$ref\":\"#/components/schemas/StatusType\"},\"path\":{\"$ref\":\"#/components/schemas/Path\"},\"url\":{\"type\":\"string\",\"title\":\"Url\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"projection\":{\"type\":\"string\",\"title\":\"Projection\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"}},\"type\":\"object\",\"title\":\"SearchItem\"},\"SearchMethod\":{\"type\":\"string\",\"enum\":[\"exact\",\"approximate\"],\"title\":\"SearchMethod\",\"description\":\"An enumeration.\"},\"SearchResult\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"URL of the asset\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"},\"metadata\":{\"$ref\":\"#/components/schemas/Metadata\"},\"in_scene_instance_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"In Scene Instance Prims\"}},\"type\":\"object\",\"required\":[\"url\",\"score\"],\"title\":\"SearchResult\",\"examples\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]},\"StatusType\":{\"enum\":[\"OK\",\"DENIED\",\"TOKEN_EXPIRED\",\"UNAUTHORIZED\",\"ES_REQUEST_ERROR\",\"ES_CONNECTION_TIMEOUT\",\"FILE_NOT_FOUND_ERROR\",\"THUMBNAIL_MISSING_ERROR\",\"INVALID_PREFIX\",\"UNKNOWN_ERROR\",\"PROJECTION_SERVICE_UNAVAILABLE\",\"REST_API_UNAVAILABLE\"],\"title\":\"StatusType\",\"description\":\"An enumeration.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}},\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T20:09:16.054Z\",\"nvcfFunctionId\":\"37435562-12fb-488b-8257-90693025981c\",\"createdDate\":\"2024-11-18T18:51:39.957Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-usdsearch\",\"blueprintUrl\":\"https://github.com/NVIDIA-Omniverse-blueprints/3d-conditioning\",\"termsOfUse\":\"GOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); use of the model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Download Blueprint\",\"url\":\"https://github.com/NVIDIA-Omniverse-blueprints/digital-twins-for-fluid-simulation\"}},\"artifactName\":\"digital-twins-for-fluid-simulation\"},\"config\":{\"name\":\"digital-twins-for-fluid-simulation\",\"type\":\"blueprint\"}}],\"items\":[\"$4c\",\"$4d\",\"$4e\",\"$4f\",\"$50\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"Develop Physical AI\"}],[\"$\",\"$L26\",null,{\"href\":\"/blueprints?filters=blueprintType%3Ablueprinttype_nvidia_omniverse\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"Pre-trained foundation models and blueprints for digital twins, synthetic data generation, and robotic simulation to accelerate Physical AI development.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"52:T1797,"])</script><script>self.__next_f.push([1,"## Use Case Overview\n\nThe NVIDIA Omniverse Blueprint for Earth-2 Weather Analytics is a reference\narchitecture, meant for independent software vendors and their application developers to\nbuild AI augmented simulation and visualization pipelines for climate and weather\ndomain.\nThis reference can be used to accelerate the development of climate tech applications\nfor weather and climate analysis, planning and risk mitigation and for developing\ndigital twins of weather and climate. \nThe blueprint showcases:\n\n* How the [NVIDIA Omniverse platform](https://www.nvidia.com/en-us/omniverse/) can be used\n to visualize geo-spatial data at scale\n* How a federated data architecture provides an adapter framework to connect partner\n platforms and open data repositories\n* How AI weather models can be deployed via [NIMs](https://www.nvidia.com/en-us/ai/) to\n produce accelerated weather forecasts\n\n## Experience Walkthrough\n\nThe blueprint experience showcases the global canvas built and rendered in Omniverse.\nVarious data streams are layered on it to provide a virtual digital twin environment for\ninteractive analysis. By integrating data from traditional weather simulation pipelines\nfrom agencies like NOAA, ECMWF with AI generated weather forecasts and observation data\nfrom providers like ESRI, Tomorrow.io, one can build next generation climate information\nsystems. \n\nIn this experience, users can pick any time frame of interest and turn on or off the\ndifferent layers to visualize global weather data like in a weather forecasting and\nanalytics application.\n\n## Architecture Diagram\n\n![Earth-2 OV Blueprint Diagram](https://assets.ngc.nvidia.com/products/api-catalog/earth2-weather-analytics/diagram.jpg)\n\n## Included NIM\n\nThe following [NIM](https://www.nvidia.com/en-us/ai/) microservices are used in this blueprint:\n\n[FourCastNet](https://build.nvidia.com/nvidia/fourcastnet) \n\n## What's Included in the Blueprint\n\nThis blueprint comes with an [open-sourced repository](https://github.com/NVIDIA-Omniverse-blueprints/earth2-weather-analytics)\nwith documentation on how to replicate and customize this blueprint.\nThe blueprint highlights the end to end pipeline that includes the following key components:\n\n* Earth-2 NIMs for generating weather forecasts using AI models\n* Data federation layer that provides the data translation interface and adapter\n mechanism to connect different data layers\n* Omniverse Kit application and extensions for rendering and streaming visualized\n geo-spatial data on a global canvas\n\n## Minimum System Requirements\n\n**Hardware Requirements**\n\n- GPU: L40S, A6000\n- RAM: 32 GB\n- CPU: x86-64 architecture, 8 Cores (Intel Core i7 (7th Generation) or AMD Ryzen 5\\)\n- Storage: 64 GB\n\n**Software Requirements**\n\n- OS: Ubuntu 22.04\n- Deployment: Kubernetes\n\n\u003e There are multiple deployment configurations that may require more compute\n\u003e requirements. For example, full deployment of the Omniverse application and NIM on the\n\u003e same machine will require 2 GPUs. For complete details, see the blueprint\n\u003e documentation on [Github](https://github.com/NVIDIA-Omniverse-blueprints/earth2-weather-analytics).\n\n## Example Walkthrough \n\nThe user interacts with the Omniverse Kit application, Earth-2 Command Center, which\noffers a high-fidelity interactive globe.\nUsing either the WebRTC APIs of the Omniverse Streaming application, or more commonly\nthe GUI of the desktop application the user can select the following:\n\n- Date-time to fetch data for (e.g. 2025-01-01 at 12:00:00 UTC)\n- Data source or AI weather model to query (e.g. Global Forecast System, GFS from NOAA)\n- Weather fields / variable (e.g. surface temperature)\n\nThis fetch request is then converted into a data federation mesh pipeline by an\nOmniverse Kit application extension that will retrieve the requested data for a set time\nframe starting at the provided date time.\nThe data federation then executes this pipeline that processes the raw weather data into\ntextures that the Omniverse application can render on the globe.\nThis data is then presented to the user as a feature layer in Omniverse that can be\nvisually customized based on use case.\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established\npolicies and practices to enable development for a wide array of AI applications.\nWhen downloaded or used in accordance with our terms of service, developers should work\nwith their supporting model team to ensure the models meet requirements for the relevant\nindustry and use case and address unforeseen product misuse. For more detailed\ninformation on ethical considerations for the models, please see the respective NIMs\nModel Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards.\nPlease report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## License\nUse of the models in this Omniverse Blueprint for Earth-2 Weather Analytics is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf). \nUse of the Omniverse SDKs and microservices is governed by the [NVIDIA Omniverse License Agreement](https://docs.omniverse.nvidia.com/enterprise/latest/common/NVIDIA_Omniverse_License_Agreement.html)\n\n## Terms of Use\nGOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf).\nUse of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/).\nThe software and materials are governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the Product-Specific Terms for [NVIDIA Omniverse](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-omniverse/)."])</script><script>self.__next_f.push([1,"53:T142f,"])</script><script>self.__next_f.push([1,"#### Experience a Real-time Wind Tunnel\n\nThis experience shows an interactive virtual wind tunnel. Simulating airflow in a virtual wind tunnel with computational fluid dynamics (CFD) requires millions of complex calculations. Without AI, it can take minutes to see the result of a single design change. Developers building the next generation of AI-powered CAE tools are combining simulation AI with immersive virtual environments to enable real-time digital twins where design changes instantly update in the simulation, as you see in this demonstration. NVIDIA is making it easier to create interactive design tools by introducing Omniverse Blueprint for interactive aerodynamics.\n\nTo enable real-time performance in a virtual wind tunnel, simulation AI models are first trained offline on representative datasets. For this demo, the training dataset was created using [Luminary Cloud’s GPU-accelerated CFD solver](https://www.luminarycloud.com/), which models complex airflow over diverse geometries. The simulation AI learns the complex relationships between car geometries (STLs) and airflow. This blueprint is compatible with industry-standard CFD solvers and can connect to third-party tools for meshing and geometry morphing, creating watertight meshes for simulation-ready geometries.\n\nThe blueprint also integrates NVIDIA PhysicsNeMo (our framework for simulation AI) with CFD solver data. This enables developers to train surrogate models from scratch or fine-tune NIM™ foundation models, reducing AI training time and cost. Once trained, the AI runs simulations orders of magnitude faster than traditional CFD, enabling real-time aerodynamic flow simulation. This speed provides designers with creative freedom, allowing designers to innovate and explore changes interactively.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/digital-twins-for-fluid-simulation/diagram.jpg)\n\n## What’s Included in the Blueprint\nNVIDIA Blueprints are comprehensive reference workflows designed to streamline AI application development across industries and accelerate deployment to production. Built with NVIDIA AI and Omniverse libraries, SDKs, and microservices, they provide a foundation for custom AI solutions. Each blueprint includes reference code for constructing workflows, tools, and documentation for deployment and customization, and a reference architecture outlining API definitions and microservice interoperability. By enabling rapid prototyping and speeding time to deployment, these blueprints empower enterprises to operationalize AI-driven solutions like AI agents, digital twins, and synthetic data generation, and more.\n\n### Included NIM Microservices\n* domino-automotive-aero 1.0\n\n### World State Controller\nThis reference solution implements an Omniverse Kit application controller that maintains the application world state and connects the 3D world stage with simulation results produced by the surrogate model.\n\nIn summary, this NVIDIA Omniverse Blueprint offers a starting point for building real-time digital twins for computer-aided engineering (CAE) workflows combining CUDA-X™ accelerated solvers, NVIDIA PhysicsNeMo for simulation AI, and Omniverse for high-quality rendering.\n\n## Minimum System Requirements\n\n#### Hardware Requirements\nThe real-time wind tunnel blueprint supports the following hardware:\n\n* At least 2x RTX™ GPUs with at least 40GB of memory each, e.g., 2xL40S or 2xA6000 \n* 128GB RAM \n* 32 CPU Cores \n* 100 GB Storage\n\n#### OS Requirements\n* Linux \\- Ubuntu 22.04 or 24.04\n\n#### Software Requirements\n* Git: For version control and repository management. \n* Git Large File System (LFS): For large files that are too large to efficiently store in a Git repository. \n* Python 3: For scripting and automation. \n* Docker: For containerized development and deployment. Ensure non-root users have Docker permissions. \n* NVIDIA Container Toolkit: For GPU-accelerated containerized development and deployment. Installation and configuring docker steps are required.\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Terms of Use\nGOVERNING TERMS: The software and materials are governed by the NVIDIA Software License Agreement (found at https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the Product-Specific Terms for NVIDIA Omniverse (found at NVIDIA Agreements | Enterprise Software | Product Specific Terms for Omniverse)."])</script><script>self.__next_f.push([1,"54:T2650,"])</script><script>self.__next_f.push([1,"This experience showcases James and Aria, our interactive digital humans who have the knowledge of NVIDIA’s products or O-RAN specifications by having direct access to corresponding knowledge bases. The Digital Human and the RAG-powered backend application use a collection of NVIDIA NIM microservices, NVIDIA ACE and Maxine technologies, and ElevenLabs text-to-speech to provide natural and immersive responses. Using James or Aria as an inspiration, users can download and customize the Digital Human for customer service blueprint for their industry use case, with document ingestion and retrieval-augmented generation (RAG), and customizing the avatar look and voice for their application. \n\n## Use Case Description\n\nThe digital human for customer service NVIDIA AI Blueprint is powered by NVIDIA Tokkio, a workflow based on ACE technologies, to bring enterprise applications to life with a 3D animated digital human interface. With approachable, human-like interactions, customer service applications can provide more engaging user experiences compared to traditional customer service options.\n\nThis workflow is designed to integrate within your existing generative AI applications built using RAG. Use this workflow to start evolving your applications running in your data center, in the cloud, or at the edge, to include a full digital human interface.\n\n## Architecture Diagram\n\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/digital-humans-for-customer-service/diagram.png)\n\n## What’s included in the Blueprint\n\n## NIM and Other Software\nThe following [NIM](https://www.nvidia.com/en-us/ai/) are used by this blueprint:\n\n* [nv-embedqa-e5-v5](https://build.nvidia.com/nvidia/nv-embedqa-e5-v5) \n* [nv-rerankqa-mistral4b-v3](https://build.nvidia.com/nvidia/nv-rerankqa-mistral-4b-v3) \n* [Llama3-8b-instruct](https://build.nvidia.com/meta/llama3-8b) \n* [Parakeet-ctc-1.1b-asr](https://build.nvidia.com/nvidia/parakeet-ctc-1\\_1b-asr)\n* [FastPitch-hifigan-tts](https://build.nvidia.com/nvidia/fastpitch-hifigan-tts) \n* [Audio2face-3D](https://build.nvidia.com/nvidia/audio2face)\n* [Audio2face-2D](https://build.nvidia.com/nvidia/audio2face-2d) \n* [Other ACE Microservices](https://developer.nvidia.com/ace)\n\nNVIDIA AI Blueprints are customizable AI workflow examples that equip enterprise developers with NIM microservices, reference code, documentation, and a Helm chart for deployment. \n\nThis [blueprint](https://github.com/NVIDIA-AI-Blueprints/digital-human) provides a reference for the users to showcase how an LLM or a RAG application can be easily connected to a digital human pipeline. The digital human and the RAG application are deployed separately. The RAG application is responsible for generating the text content of the interaction and Tokkio customer service workflow is providing a solution to enable avatar live interaction. Those two entities are separated and communicate using the REST API. The users can develop their requirements and tune the app based on their needs. Included in this workflow are steps to setup and connect both components of the customer service pipeline. Each part of the pipelines consists of the following components:\n\n**Digital Human Pipeline**\n\n* A composable Helm chart that sets up the digital human pipeline with ACE agent and deploys the Audio2Face-3D, and Riva Parakeet and FastPitch NIM microservices to deploy the default stylized avatar. The pipeline also provides different variations incorporating Audio2Face-2D to use 2D avatars.\n\n**RAG Pipeline**\n\n* A Docker Compose application deploys a Llama 3 LLM NIM, NeMo Retriever nv-embed-qa embedding NIM, a NeMo Retriever mistral-4b reranking NIM and a LangChain RAG pipeline with a FastAPI endpoint for multiturn chat. \n* Notebooks ingestion of domain specific documents [(ORAN data)](https://specifications.o-ran.org/specifications) and parameter efficient fine tuning on synthetic data generated from ORAN documents. \n\n**With this blueprint the users will be able to do the following:**\n\n1. Use the pre-built Digital Human Helm chart to create a digital human interface powered by a sample avatar asset (named Ben), Riva text-to-speech (TTS) FastPitch, and automatic speech recognition (ASR) NIM microservices. The pre-built Helm chart also connects by default to [Llama 3 8b NIM API endpoint](https://build.nvidia.com/meta/llama3-8b) to get the users’ started with an interactive Digital Human. \n2. Use the RAG application to demonstrate the power of industry knowledge with example ORAN database or ingest documents and customize the digital human knowledge for your specific industry. \n3. Able to deploy the digital human experience and the RAG application on either bare metal or their favorite cloud provider of choice with simple one-click deployment scripts.\n\n## Example Walkthrough with Sample Input/Output\n\n## **Audio2Face-3D NIM**\n\nInput \nInput Type(s): Audio \nInput Format: bytes \nInput Parameters: Tuning Parameters, Audio \nOther Properties Related to Input: Supported Sampling rates: 22.05KHz, 44.1KHz, 16KHz; All audio is resampled to 16KHz. There is no max audio length.\n\nOutput \nOutput Type(s): Blendshape Coefficients \nOutput Format: Custom Protobuf Format \nOutput Parameters: Custom Protobuf Format\n\n## **Audio2Face-2D NIM**\n\nInput Type: Portrait image, Audio\nInput Format: RGB image, 32 bit float PCM audio\nInput Parameters: 720p to 4K for the image, Audio\nOther Properties Related to Input: Supported sampling rate: 16kHz and mono channel audio. There is no max audio length.\n\n\nOutput\nOutput Format: Animated RGB Image\nOutput Parameters: Custom Protobuf Format \nOther Properties Related to Output: Input images post-processed using proprietary technique; 3 Channel, 32 bit image supported.\n\n\n## **Llama-3-8b NIM**\n\nInput\nInput Format: Text \nInput Parameters: Temperature, TopP\n\nOutput\nOutput Format: Text and code \nOutput Parameters: Max output tokens\n\n## **Riva Parakeet-ctc-1\\_1b-asr NIM** \nInput \nInput Type(s): Audio in English \nInput Format(s): Linear PCM 16-bit 1 channel\n\nOutput \nOutput Type(s): Text String in English with Timestamps\n\n## **Fastpitch-hifigan-tts NIM** \nInput\nInput Format (For FastPitch 1st Stage): Text Strings in English\nOther Properties Related to Input: 400 Character Text String Limit\nOutput \nOutput Format (For HifiGAN 2nd Stage): Audio of shape (batch x time) in wav format\n\n## **NeMo Retriever nv-embedqa-e5-v5 NIM** \nInput \nInput Type: text \nInput Format: list of strings with task-specific instructions\n\nOutput \nOutput Type: floats \nOutput Format: list of float arrays, each array containing the embeddings for the corresponding input string\n\n## **NeMo Retriever nv-rerankqa-mistral4b-v3 NIM** \nInput \nInput Type: Pair of Texts \nInput Format: List of text pairs \nOther Properties Related to Input: The model's maximum context length is 512 tokens. Texts longer than maximum length must either be chunked or truncated.\n\nOutput \nOutput Type: floats \nOutput Format: List of float arrays \nOther Properties Related to Output: Each the probability score (or raw logits) The user can decide if a Sigmoid activation function is applied to the logits.\n\nThe Audio data captured from the user is sent to ACE agent which orchestrates the communication between various NIM microservices. The ACE agent uses the Riva Parakeet NIM to convert the audio data to text which is then sent to the RAG pipeline. The RAG pipelines uses the NeMo Retriever embedding and reranking and LLM NIM microservices to answer the question with context from documents fed to it. The text result is sent to TTS, and the voice output from TTS is used to animate the digital human using the Audio2Face-3D NIM or Audio2Face-2D NIM.\n\n## API Definition \n\nAPI Interfaces for NIM collections conform to OpenAPI standards, and can be readily integrated with NVIDIA NIM containers deployed in any compatible compute cluster. Integration or replacement of API compatible components allow for easy modification of workloads to adapt to your specific use case where needed. See individual NIM [documentation](https://docs.nvidia.com/nim/index.html) for the integration details.\n\nBy default, the digital human RAG plugin has support for an API that follows the [OpenAPI specification.](https://github.com/NVIDIA-AI-Blueprints/digital-human/tree/main/api) To customize the pipeline to connect to your own RAG system, follow the instructions [here](https://github.com/NVIDIA-AI-Blueprints/digital-human).\n\n## Minimum System requirements\n\n**Hardware Requirements**\n\n**Digital human pipeline** \n\nThe digital human pipeline supports the following hardware:\n\n* T4 \n* A10 \n* L4 \n* L40S\n\nYou would need 2 GPUs minimum for 1 stream for the default 3D avatar workflow. Additional details for hardware and compute requirements for different variants of the digital human workflow can be found [here](https://docs.nvidia.com/ace/latest/workflows/tokkio/text/reference-workflows/Reference_Workflows.html).\n\n**RAG pipeline**\n\nThe RAG pipeline needs 2xA100 GPUs, one for the embedding and reranking NIM and one for the LLM NIM.\n\n**OS Requirements**\n\nBoth the digital human and the RAG pipeline can be deployed on Ubuntu 22.04 OS.\n\n\n## Terms of Use\n\nGOVERNING TERMS: \nYour use of this trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf)\nACE NIM and NGC Microservices \\- [NVIDIA AI Product License](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) \nGenerative AI Examples \\- [Apache 2](https://www.apache.org/licenses/LICENSE-2.0.txt)\\\nADDITIONAL TERMS: \nMeta Llama 3 Community License Agreement at [https://llama.meta.com/llama3/license/](https://llama.meta.com/llama3/license/)."])</script><script>self.__next_f.push([1,"55:T1f56,"])</script><script>self.__next_f.push([1,"## Use Case Description\nThe 3D conditioning for precise visual generative AI NVIDIA Omniverse Blueprint, powered by [NVIDIA NIM™](https://www.nvidia.com/en-us/ai/), [NVIDIA Omniverse](https://www.nvidia.com/en-us/omniverse/)™, [OpenUSD](https://www.nvidia.com/en-us/omniverse/usd/), image2image models such as SDXL or tuned models like Realviz4.0, and [Shutterstock Generative 3D](https://www.shutterstock.com/discover/generative-ai-3d), offers a streamlined solution for creating precise, on-brand images. This experience allows users to choose the color of a hero asset, select the desired camera angle in the 3D scene, and then use generative AI to customize scene components such as backgrounds and props. Using this experience as inspiration, developers can download and customize the blueprint to unlock use cases such as scalable concepting and ideation, through to the creation of marketing assets for their brands or customers.\n\n## Experience Walkthrough\nThe user is presented with a live 3D viewport showcasing the final product—the \"hero asset\"—created by a creative team. In this instance, the hero asset is an espresso machine with a coffee mug. This asset, representing the final design, includes all the final materials and product options. It is placed within a rudimentary scene that appears unfinished. Additional props, such as the cutting board, were generated using Shutterstock 3D Generator to populate the counter with objects unavailable to the creative team when the initial scene was created.\n\nThe user can orbit the hero asset using the left mouse button and zoom in or out with the mouse wheel. Navigation within the scene is designed to keep the hero asset in frame at all times. Once the user has identified a suitable camera angle, they can adjust the espresso machine's configuration. The machine has two control surface options, various color options, and a choice of coffee mug style. An additional control allows the user to select a pre-generated HDRi image (created with Shutterstock 360 HDRi Generator) to quickly modify the scene's background.\n\nNext, the user inputs individual prompts to generate the background. These prompts are linked to specific objects, ensuring each prompt modifies a designated area within the scene. After the user enters the prompts, the system processes them along with the scene's layout using generative AI to create the final image. During this process, the system generates masks for the targeted prompts, which the creative team can use for further image processing.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/conditioning-for-precise-visual-generative-ai/diagram.jpg)\n\n## Included NIM\nThe following [NIM](https://www.nvidia.com/en-us/ai/) are used by this blueprint: \n[USD Search](https://build.nvidia.com/nvidia/usdsearch) \n[USD Code](https://build.nvidia.com/nvidia/usdcode-llama3-70b-instruct) \n[Shutterstock 3D Generator (Playground Sample on NIM)](https://build.nvidia.com/shutterstock/edify-3d) \n[Shutterstock360 HDRi Generator (Playground Sample on NIM)](https://build.nvidia.com/shutterstock/edify-360-hdri)\n\n## What’s included in the Blueprint\n[NVIDIA Blueprints](https://nvidianews.nvidia.com/news/nvidia-and-global-partners-launch-nim-agent-blueprints-for-enterprises-to-make-their-own-ai) are customizable AI workflow examples that equip enterprise developers with NIM microservices, reference code, documentation, and a helm chart for deployment.\n\nThis blueprint provides a [reference](https://resources.nvidia.com/en-us-omniverse-product-configurator/blueprint-3d-conditioning) and [workflow guide](https://github.com/NVIDIA-Omniverse-Blueprints/3d-conditioning/tree/main) for the users to showcase how diffusion models, control nets, and corresponding auxiliary tools can be easily integrated to Omniverse to be streamed remotely. Our primary container with Omniverse handles viewport streaming and message passing between the web front-end with the second container; users can opt to use ComfyUI \\+ an image2image model such as SDXL or tuned models like Realviz4.0, leveraging our default template or take their own custom pipeline for diffusion models to handle requests coming from the first container with Omniverse. Then, we push the helm chart with the two containers. \n\n## Minimum System Requirements\nHardware Requirements\n\nGPU: 2 x L40 deployed (One for rendering the scene and another for inferencing the diffusion model) or 1x NVIDIA RTX™ 6000 Ada Generation for local\n\nCPU: x86\\_64 architecture, 8 Cores (Intel Core i7 (7th Generation) or AMD Ryzen 5\\)\n\nSystem Memory: 64GB\n\nSoftware Requirements\n\nOS: Ubuntu 22.04\n\n## Example Walkthrough with Sample Input/Output \nPrimary Container with Omniverse\n\nInput\n\nInput Type(s): JSON with payloads of text prompts and dropdown options in text\n\nInput Format: bytes\n\nOutput\n\nOutput Type(s): Viewport, Image\n\nOutput Format: stream\n\nSecond Container with Diffusion Model\n\nInput\n\nInput Type(s): JSON graph structure with embedded parameters (text, number, and image in base64) (10MB custom limit, which can be changed from the primary container)\nInput Format: bytes\n\nOutput\nOutput Type(s): Image\nOutput Format: bytes\n\nThe framework is designed to enable software developers to rapidly prototype and productize custom workflows that involve capturing buffers from the viewport of a USD scene while taking conditions in a text form, then generating an image that accounts for both constraints by running an inference of diffusion models. We have included a docker image that automatically installs and deploys ComfyUI + image2image models such as SDXL or tuned models like Realviz4.0 for the included workflow, which leverages Control Net heavily to handle multiple conditions.\n\n## Technical Considerations \nThe software is capable of conditioning a USD scene to mask a target asset to keep the 3D rendering while inferring the others via a diffusion model given the normal map or depth map as additional constraints along with text prompts. This allows users to have more artistic control over 3D assets, as well as opting to use the 3D rendering instead of the image generated by the diffusion model. To run the experience locally, you can use the downloadable and choose to run web front end or Omniverse application. Please consult our [documentation](https://github.com/NVIDIA-Omniverse-Blueprints/3d-conditioning/tree/main) to learn more about how we integrate 3D rendering and the diffusion model to achieve the final result. \n\n## Ethical Considerations\nNVIDIA believes that Trustworthy AI is a shared responsibility and we have established policies and practices to enable the development of a wide array of AI applications. When downloaded or used under our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. \n\nFor more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/). \n\n[NVIDIA Edify](https://www.nvidia.com/en-us/gpu-cloud/edify/) is a multimodal architecture for developing visual generative AI models for image, 3D, 360 HDRi, PBR materials, and video. Using NVIDIA AI Foundry, service providers can train, and customize Edify models to build commercially viable visual services on top of NVIDIA NIM. \n\n## Terms of Use\nGOVERNING TERMS: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). ADDITIONAL INFORMATION: RealvisXL license at [LICENSE.md · stabilityai/stable-diffusion-xl-base-1.0 at main.](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md)"])</script><script>self.__next_f.push([1,"56:T458,| Field | Response |\n| -- | -- |\n| Intended Application(s) \u0026 Domain(s): | Generating image embedding that is aligned with text for zero-shot classification. |\n| Model Type: | Embedding Generation |\n| Intended Users: | This model is intended for developers building search engines, classification, detection/ segmentation models. |\n| Output: | Embedding Features |\n| Describe how the model works: | This model has a vision extractor and a text encoder trained for embedding alignment |\n| Technical Limitations: | Model needs a downstream task specific head to perform CV tasks. |\n| Verified to have met prescribed NVIDIA standards: | Yes |\n| Performance Metrics: | ImageNet zero-shot accuracy |\n| Licensing: | GOVERNING TERMS: This trial is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). The use of this model is governed by the [AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/). |57:T4f1,| Field | Response |\n| -- | -- |\n| Generatable or reverse engineerable personally-identifiable information (PII)? | None |\n| Protected classes used to create this model? | Not Applicable (No PII) |\n| Was consent obtained for any personal data used? | Not Applicable (No personal data) |\n| How often is dataset reviewed? | \tBefore Release |\n| Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n| If personal data collected for the development of the model, was it collected directly by NVIDIA? |Not Applicable |\n| If personal data collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects?\t| Not Applicable |\n| If personal data collected for the development of this AI model, was it minimized to only what was required? | Not Applicable |\n| Is there provenance for all datasets used in training? | Yes |\n| Doe"])</script><script>self.__next_f.push([1,"s data labeling (annotation, metadata) comply with privacy laws? | Yes |\n| Is data compliant with data subject requests for data correction or removal, if such a request was made? | Yes |\n| Applicable NVIDIA Privacy Policy\t| [https://www.nvidia.com/en-us/about-nvidia/privacy-policy/](https://www.nvidia.com/en-us/about-nvidia/privacy-policy/) |58:T9dd,"])</script><script>self.__next_f.push([1,"**USD Search** is a versatile AI-powered search engine designed to enable comprehensive searches across images\n(e.g., .jpg, .png) and USD-based 3D models within various storage backends (AWS S3 and Omniverse Nucleus server).\nIt enables users to use natural language, image similarity, and precise metadata criteria\n(file name, type, date, size, creator, etc.) to locate relevant content efficiently. Furthermore, when integrated\nwith the Asset Graph Service, DeepSearch extends its capabilities to include searches based on USD properties and\nspatial dimensions of 3D model bounding boxes, enhancing the ability to find assets that meet specific requirements.\n\n## Features\n\n- **Natural Language Searches**: Utilize AI to search for images and USD-based 3D models using simple, descriptive\n language.\n- **Image Similarity Searches**: Find images similar to a reference image through AI-driven image comparisons.\n- **Metadata Filtering**: Filter search results by file name, file type, creation/modification dates, file size, and\n creator/modifier metadata.\n- **USD Content Filtering with Asset Graph Service**: When used with the Asset Graph Service, search capabilities are\n expanded to include filtering based on USD properties and object dimensions.\n- **Multiple Storage Backend Support**: Compatible with various storage backends, including AWS S3 bucket and Omniverse Nucleus server.\n- **Advanced File Name, Extension and Path Filters**: Use wildcards for broad or specific file name and extension searches.\n- **Date and Size Range Filtering**: Specify assets created or modified within certain date ranges or file sizes larger\n or smaller than a designated threshold.\n- **User-based Filtering**: Filter assets based on their creator or modifier, allowing for searches tailored to\n particular users' contributions.\n- **Embedding-based Similarity Threshold**: Set a similarity threshold for more nuanced control over search results in\n embedding-based searches.\n- **Custom Search Paths and Scenes**: Specify search locations within the storage backend or conduct searches within\n specific scenes for targeted results.\n- **Return Detailed Results**: Option to include images, metadata, root prims, and predictions in the search results.\n\n\nFeatures available only with the **Asset Graph Service**:\n- **USD Property Filtering**\n- **USD Object Dimension Filtering**\n- **In-scene searches**\n\n## Resources\n\n* [DeepSearch Documentation](https://docs.omniverse.nvidia.com/services/latest/services/deepsearch/overview.html)\n"])</script><script>self.__next_f.push([1,"59:T125d,"])</script><script>self.__next_f.push([1,"### Use-Case Description\n\nImitation learning lets robots learn skills from observing human demonstrations. But gathering enough high-quality real-world datasets can be challenging, costly, and time-consuming. Synthetic data, generated from physically accurate simulations, addresses the challenge of limited real-world data acquisition by accelerating data collection and providing the diversity needed to generalize robot learning models.\n\nThe NVIDIA Isaac GR00T blueprint for synthetic manipulation motion generation is the ideal place to start. This is a reference workflow for creating exponentially large amounts of synthetic motion trajectories for robot manipulation from a small number of human demonstrations, built on NVIDIA Omniverse™ and NVIDIA Cosmos™.\n\nFirst, developers use a spatial computing device such as the Apple Vision Pro to portal into their simulated robot digital twin and record motion demonstration teleoperating a simulated robot. These recordings are then used to generate a larger set of physically accurate synthetic motion trajectories. Finally, the blueprint further augments the dataset by generating an exponentially large, photorealistic, and diverse set of training data. \n\n\u003cdiv style=\"background-color: #202020; padding: 16px; border-radius: 16px; color: #fff; line-height: 21px;\"\u003e\nNote: The first release of this blueprint is for single-arm manipulation only. Support for bi-manual humanoid robot manipulation is coming soon.\n\u003c/div\u003e\n\n### Experience Walkthrough\n\nThe overall experience is divided into four distinct parts: \n\n1\\. Choose from a pre-recorded set of human demonstrations.\n\n2\\. View the synthetically generated motion.\n\n3\\. Select from the list of pre-populated prompts to augment the generated motions.\n\n4\\. Click \"View Source Code” to retrieve the blueprint from GitHub.\n\n## Architecture Diagram\n![](https://assets.ngc.nvidia.com/products/api-catalog/gr00t/diagram.jpg)\n\n## What’s Included in the Blueprint\n\n##### Sample Recorded Data\n\n* Pre-recorded human demonstrations for a single-arm manipulation\n\n##### Robot Simulation and Training Frameworks\n\n* NVIDIA Isaac™ Lab, an open-source, unified framework for [robot learning](https://www.nvidia.com/en-us/glossary/robot-learning/) designed to help train robot policies built on Isaac Sim\n\n##### Data Generation\n\n* GR00T-Mimic, a feature in Isaac Lab, uses the recorded demonstrations as input to generate synthetic motion trajectories \n\n##### Data Augmentation\n\n* GR00T-Gen, a feature in Isaac Lab for augmenting 3D datasets to achieve the necessary photorealism and diversity \n* [Cosmos-Transfer1-7B](https://huggingface.co/nvidia/Cosmos-Transfer1-7B) model\n\n### File Deliverables\n\n**Input:** \n\n- Pre-selected collection of human demonstration recordings, captured with teleoperation in simulation \n- Pre-populated prompts to augment data with prepopulated prompts\n\n**Output:** \n\n- Synthetically-generated trajectories \n- Augmented video displayed on the screen\n- Jupyter notebook to recreate the end-to-end development experience\n \n#### Minimum System Requirements\nHardware Requirements\n\nGPU\n* NVIDIA 6000 Ada, 4090, 5090, L40, L40S, L20 and A40 or any higher level NVIDIA RTX™-capable GPU \n* Cosmos - HGX node (1x H100) TBD\n\nCPU\n* Intel Core i7 (7th Generation) \n* AMD Ryzen 5\n\n#### OS Requirements\n* Ubuntu 22.04 OS \n* Windows 11\n\n## Ethical Considerations\nNVIDIA believes trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting team to ensure the technologies meet requirements for the relevant industry and use case and address unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Licenses\n \nLicensing information for Isaac Lab can be found [here](https://isaac-sim.github.io/IsaacLab/main/source/refs/license.html).\n\nLicense information for For NVIDIA Cosmos: can be found under [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\n\nLicensing for GR00T-Mimic can be found [here](https://www.google.com/url?q=https://github.com/NVIDIA-Omniverse-blueprints/synthetic-manipulation-motion-generation\u0026sa=D\u0026source=docs\u0026ust=1741721959187379\u0026usg=AOvVaw2ACeuFe6HOCLbIpMpcnrKH)\n\n\n## Terms of Use\nGoverning Terms: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf)."])</script><script>self.__next_f.push([1,"5a:T13da,"])</script><script>self.__next_f.push([1,"## Use Case Description\n\n[Physical AI](https://www.nvidia.com/en-us/glossary/generative-physical-ai/) is poised to transform manufacturing, supply chain and logistics—bringing unprecedented levels of industrial automation, intelligence, and autonomy to the world’s factories, warehouses, and industrial facilities.\n\nIn the smart factories and warehouses of today and of the future, humans and fleets of robots, including AGVs/AMRs, [humanoid robots](https://www.nvidia.com/en-us/use-cases/humanoid-robots/), intelligent cameras, and [visual AI agents](https://www.nvidia.com/en-us/autonomous-machines/intelligent-video-analytics-platform/) work together to achieve their objectives. To ensure their efficient operation in the real world, enterprises will rely on [digital twins](https://www.nvidia.com/en-us/glossary/digital-twin/) of their facilities to simulate interactions and performance of these different robot types and their objectives, ensuring they can work together seamlessly to accomplish their tasks.\n\nThe Mega NVIDIA Omniverse Blueprint, powered by [NVIDIA Omniverse™](https://www.nvidia.com/en-us/omniverse/), [OpenUSD](https://www.nvidia.com/en-us/omniverse/usd/), and [Isaac™ ROS](https://developer.nvidia.com/isaac/ros), enables enterprises to combine real-time [sensor simulation](https://www.nvidia.com/en-us/glossary/sensor-simulation/) and [synthetic data generation](https://www.nvidia.com/en-us/use-cases/synthetic-data/) to simulate these complex human-robot interactions and verify the performance of physical AI systems in [industrial digital twins](https://www.nvidia.com/en-us/use-cases/ai-for-virtual-factory-solutions/?deeplink=content-tab--1) before real-world deployment.\n\n## Experience Walkthrough\n\nWhen starting the experience, users are presented with a sample warehouse populated with racks, boxes, and autonomous mobile robots (AMRs) equipped with 3D LiDAR and RGB camera sensors.\n\nTo set up the simulation, users can select a location inside the warehouse and configure two AMRs. For each AMR configuration, users can:\n\n- Select either a “smart” robot that can detect and avoid obstacles on its path or a “simple” robot that can only follow preprogrammed paths. \n- Select from one of the four camera views (front, left, right, back) and the type of render they want to generate. \n- Create an AMR path to navigate interactively. \n\nOnce the AMRs are configured, the user clicks the “Run Simulation” button, and the AMR brain, World Simulator, and Sensor RTX™ service are deployed. As shown in the architectural diagram:\n\n- The AMR brains control the AMRs and send control signals to actuate them in the World Simulator. \n- The World Simulator runs the physics-based simulation of the AMR’s movement. \n- Each AMR has multiple sensors that are simulated: one 3D LiDAR, one IMU, and one RGB camera with multiple [AOVs](https://www.nvidia.com/en-us/on-demand/session/omniverse2020-om1458/), in addition to a top-view camera using NVIDIA Sensor RTX APIs. \n- Sensor data is streamed back to the AMR controllers to perceive the surroundings and determine the next step of control signals. \n\nNote that simulations typically take 15–20 minutes to complete. During periods of high demand, results may take longer to generate. Users receive a simulation ID (valid for 14 days) that allows them to return to the experience to view the results. While waiting for the simulation to complete, users receive periodic simulation progress updates and an [introduction video to the reference architecture](https://assets.ngc.nvidia.com/products/api-catalog/mega/explanation.mp4). To get more in-depth information about the blueprint, [read the technical blog](https://developer.nvidia.com/blog/simulating-robots-in-industrial-facility-digital-twins/).\n\n## Architecture Diagram\n![](https://assets.ngc.nvidia.com/products/api-catalog/mega/diagram.jpg)\n\n## What’s Included in the Blueprint\n\nNVIDIA Blueprints are comprehensive reference workflows designed to streamline AI application development across industries and accelerate deployment to production. Built with NVIDIA AI and Omniverse libraries, SDKs, and microservices, they provide a foundation for custom AI solutions. Each blueprint includes a reference code for constructing workflows, tools and documentation for deployment and customization, and a reference architecture outlining API definitions and microservice interoperability. By enabling rapid prototyping and speeding time to deployment, these blueprints empower enterprises to operationalize AI-driven solutions like AI agents, digital twins, synthetic data generation, and more.\n\n## Terms of Use\nGOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); use of the model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/)."])</script><script>self.__next_f.push([1,"5b:{\"name\":\"earth2-weather-analytics\",\"type\":\"blueprint\"}\n5c:{\"name\":\"digital-twins-for-fluid-simulation\",\"type\":\"blueprint\"}\n5d:{\"name\":\"digital-humans-for-customer-service\",\"type\":\"blueprint\"}\n5e:{\"name\":\"conditioning-for-precise-visual-generative-ai\",\"type\":\"blueprint\"}\n5f:{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"type\":\"blueprint\"}\n60:{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"type\":\"blueprint\"}\n"])</script><script>self.__next_f.push([1,"3b:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"earth2-weather-analytics\",\"displayName\":\"AI Weather Analytics with Earth-2\",\"publisher\":\"nvidia\",\"shortDescription\":\"Develop AI powered weather analysis and forecasting application visualizing multi-layered geospatial data.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/earth2-weather-analytics.jpg\",\"labels\":[\"AI Weather Prediction\",\"Blueprint\",\"Climate Science\",\"Earth-2\",\"NVIDIA AI\",\"Weather Simulation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"NIM\",\"value\":\"fourcastnet\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:22:00.681Z\",\"description\":\"$52\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:22:00.681Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"29ea023f-e2fd-419f-8dd4-a5511507d2c8\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"nvcfFunctionId\":\"f7b349b8-f3e7-4636-8802-4585bf5c99ab\",\"createdDate\":\"2025-03-18T19:22:01.013Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"GOVERNING TERMS: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\"\u003eNVIDIA Community Model License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-Omniverse-blueprints/earth2-weather-analytics\"}},\"artifactName\":\"earth2-weather-analytics\"},\"config\":{\"name\":\"earth2-weather-analytics\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"digital-twins-for-fluid-simulation\",\"displayName\":\"Build a Digital Twin for Interactive Fluid Simulation\",\"publisher\":\"nvidia\",\"shortDescription\":\"This NVIDIA Omniverse™ Blueprint demonstrates how commercial software vendors can create interactive digital twins.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/digital-twins-for-fluid-simulation.jpg\",\"labels\":[\"Blueprint\",\"CAE\",\"Computer-aided-engineering\",\"External Aerodynamics\",\"NVIDIA Omniverse\",\"simulation\"],\"attributes\":[{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-11-18T18:51:39.577Z\",\"description\":\"$53\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T20:09:15.623Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"31cd111e-f34f-4870-928b-b6f927bff889\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Build a Digital Twin for Interactive Fluid Simulation\",\"description\":\"TBD\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"Name\",\"url\":\"https://place\"}},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"/nvidia/digital-twins-for-fluid-simulation\":{\"post\":{\"summary\":\"Search Post\",\"operationId\":\"search_post_nvidia_usdcava_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/DeepSearchSearchRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Search results\",\"content\":{\"application/json\":{\"schema\":{\"items\":{\"$ref\":\"#/components/schemas/SearchResult\"},\"type\":\"array\",\"title\":\"Response Search Post V2 USD Search Post\"},\"example\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"402\":{\"description\":\"Payment Required\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/TBD\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}},\"x-nvai-meta\":{\"templates\":[{\"requestEjs\":{\"python\":\"import requests\\nimport json\\n\\nresponse = requests.post(\\n url='https://ai.api.nvidia.com/v1/nvidia/digital-twins-for-fluid-simulation', \\n headers={\\n 'Authorization': 'Bearer $NVIDIA_API_KEY',\\n 'Accept': 'application/json',\\n 'Content-Type': 'application/json',\\n },\\n data=json.dumps(\\n dict(\\n description = 'pallet', \\n file_extension_include ='usd*', \\n return_images ='true', \\n return_metadata = 'true', \\n return_vision_generated_metadata = 'true', \\n cutoff_threshold = '1.05', \\n limit = '50'\\n )\\n )\\n)\\n\\nprint(response.status_code)\\nprint(json.dumps(response.json(), indent=2))\",\"shell\":\"curl --request POST --url 'https://ai.api.nvidia.com/v1/nvidia/digital-twins-for-fluid-simulation' --header 'Authorization: Bearer $NVIDIA_API_KEY' --header 'Accept: application/json' --header 'Content-Type: application/json' --data '\\n{\\n 'description': 'pallet',\\n 'file_extension_include': 'usd*',\\n 'return_images': 'true',\\n 'return_metadata': 'true',\\n 'return_vision_generated_metadata': 'true',\\n 'cutoff_threshold': '1.05',\\n 'limit': '50'\\n}'\\n\"}}]}}},\"security\":[{\"Token\":[]}],\"components\":{\"schemas\":{\"DeepSearchSearchRequest\":{\"properties\":{\"description\":{\"type\":\"string\",\"title\":\"Description\",\"description\":\"Conduct text-based searches powered by AI\",\"default\":\"\"},\"image_similarity_search\":{\"type\":\"string\",\"title\":\"Image Similarity Search\",\"description\":\"Perform similarity searches based on images\",\"default\":\"\"},\"file_name\":{\"type\":\"string\",\"title\":\"File Name\",\"description\":\"Filter results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"exclude_file_name\":{\"type\":\"string\",\"title\":\"Exclude File Name\",\"description\":\"Exclude results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_include\":{\"type\":\"string\",\"title\":\"File Extension Include\",\"description\":\"Filter results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_exclude\":{\"type\":\"string\",\"title\":\"File Extension Exclude\",\"description\":\"Exclude results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"created_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created After\",\"description\":\"Filter results to only include assets created after a specified date\",\"default\":\"\"},\"created_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created Before\",\"description\":\"Filter results to only include assets created before a specified date\",\"default\":\"\"},\"modified_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified After\",\"description\":\"Filter results to only include assets modified after a specified date\",\"default\":\"\"},\"modified_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified Before\",\"description\":\"Filter results to only include assets modified before a specified date\",\"default\":\"\"},\"file_size_greater_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Greater Than\",\"description\":\"Filter results to only include files larger than a specific size\",\"default\":\"\"},\"file_size_less_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Less Than\",\"description\":\"Filter results to only include files smaller than a specific size\",\"default\":\"\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\",\"description\":\"Filter results to only include assets created by a specific user. In case AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_created_by\":{\"type\":\"string\",\"title\":\"Exclude Created By\",\"description\":\"Exclude assets created by a specific user from the results\",\"default\":\"\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\",\"description\":\"Filter results to only include assets modified by a specific user. In the case, when AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_modified_by\":{\"type\":\"string\",\"title\":\"Exclude Modified By\",\"description\":\"Exclude assets modified by a specific user from the results\",\"default\":\"\"},\"similarity_threshold\":{\"type\":\"number\",\"maximum\":2,\"minimum\":0,\"title\":\"Similarity Threshold\",\"description\":\"Set the similarity threshold for embedding-based searches. This functionality allows filterring duplicates and returning only those results that are different from each other. Assets are considered to be duplicates if the cosine distance betwen the embeddings a smaller than the similarity_threshold value, which could be in the [0, 2] range.\"},\"cutoff_threshold\":{\"type\":\"number\",\"minimum\":0,\"title\":\"Cutoff Threshold\",\"description\":\"Set the cutoff threshold for embedding-based searches\"},\"search_path\":{\"type\":\"string\",\"title\":\"Search Path\",\"description\":\"Specify the search path within the storage backend. This path should not contain the storage backend URL, just the asset path on the storage backend. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"search_in_scene\":{\"type\":\"string\",\"title\":\"Search In Scene\",\"description\":\"Conduct the search within a specific scene. Provide the full URL for the asset including the storage backend URL prefix.\",\"default\":\"\"},\"filter_by_properties\":{\"type\":\"string\",\"title\":\"Filter By Properties\",\"description\":\"Filter assets by USD attributes where at least one root prim matches (note: only supported for a subset of attributes indexed). Format: `attribute1=abc,attribute2=456`\",\"default\":\"\"},\"min_bbox_x\":{\"type\":\"number\",\"title\":\"Min Bbox X\",\"description\":\"Filter by minimum X axis dimension of the asset's bounding box\"},\"min_bbox_y\":{\"type\":\"number\",\"title\":\"Min Bbox Y\",\"description\":\"Filter by minimum Y axis dimension of the asset's bounding box\"},\"min_bbox_z\":{\"type\":\"number\",\"title\":\"Min Bbox Z\",\"description\":\"Filter by minimum Z axis dimension of the asset's bounding box\"},\"max_bbox_x\":{\"type\":\"number\",\"title\":\"Max Bbox X\",\"description\":\"Filter by maximum X axis dimension of the asset's bounding box\"},\"max_bbox_y\":{\"type\":\"number\",\"title\":\"Max Bbox Y\",\"description\":\"Filter by maximum Y axis dimension of the asset's bounding box\"},\"max_bbox_z\":{\"type\":\"number\",\"title\":\"Max Bbox Z\",\"description\":\"Filter by maximum Z axis dimension of the asset's bounding box\"},\"return_images\":{\"type\":\"boolean\",\"title\":\"Return Images\",\"description\":\"Return images if set to True\",\"default\":false},\"return_metadata\":{\"type\":\"boolean\",\"title\":\"Return Metadata\",\"description\":\"Return metadata if set to True\",\"default\":false},\"return_root_prims\":{\"type\":\"boolean\",\"title\":\"Return Root Prims\",\"description\":\"Return root prims if set to True\",\"default\":false},\"return_predictions\":{\"type\":\"boolean\",\"title\":\"Return Predictions\",\"description\":\"Return predictions if set to True\",\"default\":false},\"return_in_scene_instances_prims\":{\"type\":\"boolean\",\"title\":\"Return In Scene Instances Prims\",\"description\":\"[in-scene search only] Return prims of instances of objects found in the scene\",\"default\":false},\"embedding_knn_search_method\":{\"allOf\":[{\"$ref\":\"#/components/schemas/SearchMethod\"}],\"description\":\"Search method, approximate should be faster but is less accurate. Default is exact\"},\"limit\":{\"type\":\"integer\",\"maximum\":10000,\"minimum\":1,\"title\":\"Limit\",\"description\":\"Set the maximum number of results to return from the search, default is 32\"},\"vision_metadata\":{\"type\":\"string\",\"title\":\"Vision Metadata\",\"description\":\"Input string for the vision metadata query\"},\"return_vision_generated_metadata\":{\"type\":\"boolean\",\"title\":\"Return Vision Generated Metadata\",\"description\":\"Return vision generated metadata if True\",\"default\":false}},\"type\":\"object\",\"title\":\"DeepSearchSearchRequest\",\"example\":{\"description\":\"box\",\"return_metadata\":true,\"limit\":16}},\"EdgeType\":{\"type\":\"string\",\"enum\":[\"depends_on\",\"parent_prim\",\"root_prim\",\"source_asset\"],\"title\":\"EdgeType\",\"description\":\"An enumeration.\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"minimum\":100,\"maximum\":999,\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"format\":\"^.{1, 1024}$\",\"maxLength\":1024,\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"format\":\"^.{1, 256}$\",\"maxLength\":256,\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36,\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"PaymentRequiredError\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"},\"HealthResponse\":{\"properties\":{\"health_code\":{\"type\":\"string\",\"title\":\"Health Code\",\"description\":\"Status code, OK if successful, otherwise error code.\",\"examples\":[\"200\"]},\"health_code_description\":{\"type\":\"string\",\"title\":\"Health Code Description\",\"description\":\"Status description.\",\"examples\":[\"OK\"]},\"headers\":{\"type\":\"object\",\"title\":\"Headers\",\"description\":\"Response headers.\",\"default\":{\"cache-control\":[\"no-cache\",\"no-store\",\"max-age=0\",\"must-revalidate\"],\"expires\":0,\"pragma\":\"no-cache\"}}},\"type\":\"object\",\"required\":[\"health_code\",\"health_code_description\"],\"title\":\"HealthResponse\"},\"Metadata\":{\"properties\":{\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"}},\"type\":\"object\",\"title\":\"Metadata\"},\"Path\":{\"properties\":{\"version\":{\"type\":\"integer\",\"title\":\"Version\"},\"type\":{\"$ref\":\"#/components/schemas/PathType\"},\"uri\":{\"type\":\"string\",\"title\":\"Uri\"},\"acl\":{\"items\":{\"$ref\":\"#/components/schemas/PathPermission\"},\"type\":\"array\"},\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"},\"event\":{\"$ref\":\"#/components/schemas/PathEvent\"},\"mounted\":{\"type\":\"boolean\",\"title\":\"Mounted\"},\"transaction_id\":{\"type\":\"string\",\"title\":\"Transaction Id\"},\"destination\":{\"type\":\"string\",\"title\":\"Destination\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"is_deleted\":{\"type\":\"boolean\",\"title\":\"Is Deleted\"},\"deleted_by\":{\"type\":\"string\",\"title\":\"Deleted By\"},\"deleted_timestamp\":{\"type\":\"number\",\"title\":\"Deleted Timestamp\"}},\"type\":\"object\",\"title\":\"Path\"},\"PathEvent\":{\"enum\":[\"full\",\"create\",\"update\",\"delete\",\"change_acl\",\"set_path_options\",\"lock\",\"unlock\",\"rename\",\"copy\",\"replace_version\"],\"title\":\"PathEvent\",\"description\":\"An enumeration.\"},\"PathPermission\":{\"enum\":[\"read\",\"write\",\"admin\"],\"title\":\"PathPermission\",\"description\":\"An enumeration.\"},\"PathType\":{\"enum\":[\"\",\"asset\",\"folder\",\"channel\",\"mount\",\"object\",\"none\"],\"title\":\"PathType\",\"description\":\"An enumeration.\"},\"Prediction\":{\"properties\":{\"tag\":{\"type\":\"string\",\"title\":\"Tag\"},\"prob\":{\"type\":\"number\",\"title\":\"Prob\"}},\"type\":\"object\",\"required\":[\"tag\",\"prob\"],\"title\":\"Prediction\"},\"Prim\":{\"properties\":{\"scene_url\":{\"type\":\"string\",\"title\":\"Scene Url\"},\"usd_path\":{\"type\":\"string\",\"title\":\"Usd Path\"},\"prim_type\":{\"type\":\"string\",\"title\":\"Prim Type\"},\"bbox_max\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Max\"},\"bbox_min\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Min\"},\"bbox_midpoint\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Midpoint\"},\"bbox_dimension_x\":{\"type\":\"number\",\"title\":\"Bbox Dimension X\"},\"bbox_dimension_y\":{\"type\":\"number\",\"title\":\"Bbox Dimension Y\"},\"bbox_dimension_z\":{\"type\":\"number\",\"title\":\"Bbox Dimension Z\"},\"source_asset_url\":{\"type\":\"string\",\"title\":\"Source Asset Url\"},\"properties\":{\"type\":\"object\",\"title\":\"Properties\"}},\"type\":\"object\",\"required\":[\"scene_url\",\"usd_path\",\"prim_type\",\"bbox_max\",\"bbox_min\",\"bbox_midpoint\",\"bbox_dimension_x\",\"bbox_dimension_y\",\"bbox_dimension_z\"],\"title\":\"Prim\"},\"SearchItem\":{\"properties\":{\"status\":{\"$ref\":\"#/components/schemas/StatusType\"},\"path\":{\"$ref\":\"#/components/schemas/Path\"},\"url\":{\"type\":\"string\",\"title\":\"Url\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"projection\":{\"type\":\"string\",\"title\":\"Projection\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"}},\"type\":\"object\",\"title\":\"SearchItem\"},\"SearchMethod\":{\"type\":\"string\",\"enum\":[\"exact\",\"approximate\"],\"title\":\"SearchMethod\",\"description\":\"An enumeration.\"},\"SearchResult\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"URL of the asset\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"},\"metadata\":{\"$ref\":\"#/components/schemas/Metadata\"},\"in_scene_instance_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"In Scene Instance Prims\"}},\"type\":\"object\",\"required\":[\"url\",\"score\"],\"title\":\"SearchResult\",\"examples\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]},\"StatusType\":{\"enum\":[\"OK\",\"DENIED\",\"TOKEN_EXPIRED\",\"UNAUTHORIZED\",\"ES_REQUEST_ERROR\",\"ES_CONNECTION_TIMEOUT\",\"FILE_NOT_FOUND_ERROR\",\"THUMBNAIL_MISSING_ERROR\",\"INVALID_PREFIX\",\"UNKNOWN_ERROR\",\"PROJECTION_SERVICE_UNAVAILABLE\",\"REST_API_UNAVAILABLE\"],\"title\":\"StatusType\",\"description\":\"An enumeration.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}},\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T20:09:16.054Z\",\"nvcfFunctionId\":\"37435562-12fb-488b-8257-90693025981c\",\"createdDate\":\"2024-11-18T18:51:39.957Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-usdsearch\",\"blueprintUrl\":\"https://github.com/NVIDIA-Omniverse-blueprints/3d-conditioning\",\"termsOfUse\":\"GOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); use of the model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Download Blueprint\",\"url\":\"https://github.com/NVIDIA-Omniverse-blueprints/digital-twins-for-fluid-simulation\"}},\"artifactName\":\"digital-twins-for-fluid-simulation\"},\"config\":{\"name\":\"digital-twins-for-fluid-simulation\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"digital-humans-for-customer-service\",\"displayName\":\"Build a Digital Human\",\"publisher\":\"nvidia\",\"shortDescription\":\"Create intelligent, interactive avatars for customer service across industries\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/digital-humans-for-customer-service.jpg\",\"labels\":[\"Audio-to-face\",\"Blueprint\",\"Chat\",\"Digital Humans\",\"Speech-to-text\",\"NVIDIA AI\",\"NVIDIA Omniverse\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"NIM\",\"value\":\"audio2face-3d,llama3-8b,parakeet-ctc-1_1b-asr\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-08-26T20:28:47.307Z\",\"description\":\"$54\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:36:10.618Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"8a253105-cc23-4f0b-837d-6680fdec9f43\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:36:11.110Z\",\"nvcfFunctionId\":\"1feccd47-0f9a-48c6-85e6-0647d8d3ab3e\",\"createdDate\":\"2024-08-26T20:28:47.559Z\",\"attributes\":{\"apiDocsUrl\":null,\"blueprintUrl\":\"https://github.com/NVIDIA-AI-Blueprints/digital-human\",\"termsOfUse\":\"\",\"showUnavailableBanner\":false,\"avatars\":[{\"function_id\":\"41e5a001-b4f2-48ab-b834-8042176277a6\",\"name\":\"james\",\"title\":\"James - 3D Character\",\"description\":\"Meet James, interactive digital human, a communications expert powered by the selected knowledge base \u0026 animated by Audio2Face-3D.\\n\",\"disabled\":false},{\"function_id\":\"fe0f0bc1-f649-4842-93f1-121344b5413f\",\"name\":\"aria\",\"title\":\"Aria - 2D Character\",\"description\":\"Meet Aria, an interactive digital human, an analyst with expertise in selected knowledge base \u0026 animated by Audio2Face-2D.\\n\",\"disabled\":false}],\"knowledge_bases\":[{\"id\":\"nvidia-rag-chain-server:8090\",\"title\":\"NVIDIA Product documentation\",\"description\":\"This Knowledge base contains documents about NVIDIA products.\\n\",\"disabled\":false},{\"id\":\"telco-rag-chain-server:8091\",\"title\":\"Telco ORAN specifications\",\"description\":\"This knowledge base contains O-RAN specifications that aims to promote interoperability and openness in telecommunication standards.\\n\",\"disabled\":false}],\"iframeUrl\":\"https://nvcf-ifrm-prod.tokkio-nvcf.nvidia.com/nvPlayer.html\",\"nimTabDescription\":\"These NVIDIA NIM are used to create James, in addition to ElevenLabs' text-to-speech and several other NVIDIA ACE microservices.\"},\"artifactName\":\"digital-humans-for-customer-service\"},\"config\":{\"name\":\"digital-humans-for-customer-service\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"conditioning-for-precise-visual-generative-ai\",\"displayName\":\"3D Conditioning for Precise Visual Generative AI\",\"publisher\":\"nvidia\",\"shortDescription\":\"Enhance and modify high-quality compositions using real-time rendering and generative AI output without affecting a hero product asset.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/conditioning-for-precise-visual-generative-ai.jpg\",\"labels\":[\"Blueprint\",\"NVIDIA Omniverse\",\"simulation\",\"visual design\"],\"attributes\":[{\"key\":\"NIM\",\"value\":\"usdcode,usdsearch,edify-3d,edify-360-hdri\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"| Field | Response |\\n| -- | -- |\\n|Participation considerations from adversely impacted groups [(protected classes)](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None of the Above |\\n| Measures taken to mitigate against unwanted bias: | Scaled Dataset and adding augmentations.|\",\"canGuestDownload\":true,\"createdDate\":\"2024-11-04T17:12:32.415Z\",\"description\":\"$55\",\"explainability\":\"$56\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$57\",\"safetyAndSecurity\":\"| Field | Response |\\n| -- | -- |\\n| Model Application(s): | Embedding generation and retrieval for OpenUSD assets |\\n| Describe the life-critical application (if present). | None: Not within Operational Design Domain |\\n| Use Case Restrictions: | Abide by [https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) |\\n| Describe access restrictions (if any): | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. |\",\"updatedDate\":\"2025-02-07T16:39:15.641Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"5c169559-fcfa-4f2b-bd9f-4ea1154f51af\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"USD Search API\",\"description\":\"$58\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"Name\",\"url\":\"https://place\"}},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"/nvidia/conditioning-for-precise-visual-generative-ai\":{\"post\":{\"summary\":\"Search Post\",\"operationId\":\"search_post_nvidia_usdcava_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/DeepSearchSearchRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Search results\",\"content\":{\"application/json\":{\"schema\":{\"items\":{\"$ref\":\"#/components/schemas/SearchResult\"},\"type\":\"array\",\"title\":\"Response Search Post V2 USD Search Post\"},\"example\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"402\":{\"description\":\"Payment Required\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/TBD\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}},\"x-nvai-meta\":{\"templates\":[{\"requestEjs\":{\"python\":\"import requests\\nimport json\\n\\nresponse = requests.post(\\n url='https://ai.api.nvidia.com/v1/nvidia/conditioning-for-precise-visual-generative-ai', \\n headers={\\n 'Authorization': 'Bearer $NVIDIA_API_KEY',\\n 'Accept': 'application/json',\\n 'Content-Type': 'application/json',\\n },\\n data=json.dumps(\\n dict(\\n description = 'pallet', \\n file_extension_include ='usd*', \\n return_images ='true', \\n return_metadata = 'true', \\n return_vision_generated_metadata = 'true', \\n cutoff_threshold = '1.05', \\n limit = '50'\\n )\\n )\\n)\\n\\nprint(response.status_code)\\nprint(json.dumps(response.json(), indent=2))\",\"shell\":\"curl --request POST --url 'https://ai.api.nvidia.com/v1/nvidia/conditioning-for-precise-visual-generative-ai' --header 'Authorization: Bearer $NVIDIA_API_KEY' --header 'Accept: application/json' --header 'Content-Type: application/json' --data '\\n{\\n 'description': 'pallet',\\n 'file_extension_include': 'usd*',\\n 'return_images': 'true',\\n 'return_metadata': 'true',\\n 'return_vision_generated_metadata': 'true',\\n 'cutoff_threshold': '1.05',\\n 'limit': '50'\\n}'\\n\"}}]}}},\"security\":[{\"Token\":[]}],\"components\":{\"schemas\":{\"DeepSearchSearchRequest\":{\"properties\":{\"description\":{\"type\":\"string\",\"title\":\"Description\",\"description\":\"Conduct text-based searches powered by AI\",\"default\":\"\"},\"image_similarity_search\":{\"type\":\"string\",\"title\":\"Image Similarity Search\",\"description\":\"Perform similarity searches based on images\",\"default\":\"\"},\"file_name\":{\"type\":\"string\",\"title\":\"File Name\",\"description\":\"Filter results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"exclude_file_name\":{\"type\":\"string\",\"title\":\"Exclude File Name\",\"description\":\"Exclude results by asset file name, allowing partial matches. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_include\":{\"type\":\"string\",\"title\":\"File Extension Include\",\"description\":\"Filter results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"file_extension_exclude\":{\"type\":\"string\",\"title\":\"File Extension Exclude\",\"description\":\"Exclude results by file extension. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"created_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created After\",\"description\":\"Filter results to only include assets created after a specified date\",\"default\":\"\"},\"created_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Created Before\",\"description\":\"Filter results to only include assets created before a specified date\",\"default\":\"\"},\"modified_after\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified After\",\"description\":\"Filter results to only include assets modified after a specified date\",\"default\":\"\"},\"modified_before\":{\"type\":\"string\",\"pattern\":\"\\\\d{4}-\\\\d{2}-\\\\d{2}\",\"title\":\"Modified Before\",\"description\":\"Filter results to only include assets modified before a specified date\",\"default\":\"\"},\"file_size_greater_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Greater Than\",\"description\":\"Filter results to only include files larger than a specific size\",\"default\":\"\"},\"file_size_less_than\":{\"type\":\"string\",\"pattern\":\"\\\\d+[KMGT]B\",\"title\":\"File Size Less Than\",\"description\":\"Filter results to only include files smaller than a specific size\",\"default\":\"\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\",\"description\":\"Filter results to only include assets created by a specific user. In case AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_created_by\":{\"type\":\"string\",\"title\":\"Exclude Created By\",\"description\":\"Exclude assets created by a specific user from the results\",\"default\":\"\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\",\"description\":\"Filter results to only include assets modified by a specific user. In the case, when AWS S3 bucket is used as a storage backend, this field corresponds to the owner's ID. In case of an Omniverse Nucleus server, this field may depend on the configuration, but typically corresponds to user email.\",\"default\":\"\"},\"exclude_modified_by\":{\"type\":\"string\",\"title\":\"Exclude Modified By\",\"description\":\"Exclude assets modified by a specific user from the results\",\"default\":\"\"},\"similarity_threshold\":{\"type\":\"number\",\"maximum\":2,\"minimum\":0,\"title\":\"Similarity Threshold\",\"description\":\"Set the similarity threshold for embedding-based searches. This functionality allows filterring duplicates and returning only those results that are different from each other. Assets are considered to be duplicates if the cosine distance betwen the embeddings a smaller than the similarity_threshold value, which could be in the [0, 2] range.\"},\"cutoff_threshold\":{\"type\":\"number\",\"minimum\":0,\"title\":\"Cutoff Threshold\",\"description\":\"Set the cutoff threshold for embedding-based searches\"},\"search_path\":{\"type\":\"string\",\"title\":\"Search Path\",\"description\":\"Specify the search path within the storage backend. This path should not contain the storage backend URL, just the asset path on the storage backend. Use wildcards: `*` for any number of characters, `?` for a single character. Separate terms with `,` for OR and `;` for AND.\",\"default\":\"\"},\"search_in_scene\":{\"type\":\"string\",\"title\":\"Search In Scene\",\"description\":\"Conduct the search within a specific scene. Provide the full URL for the asset including the storage backend URL prefix.\",\"default\":\"\"},\"filter_by_properties\":{\"type\":\"string\",\"title\":\"Filter By Properties\",\"description\":\"Filter assets by USD attributes where at least one root prim matches (note: only supported for a subset of attributes indexed). Format: `attribute1=abc,attribute2=456`\",\"default\":\"\"},\"min_bbox_x\":{\"type\":\"number\",\"title\":\"Min Bbox X\",\"description\":\"Filter by minimum X axis dimension of the asset's bounding box\"},\"min_bbox_y\":{\"type\":\"number\",\"title\":\"Min Bbox Y\",\"description\":\"Filter by minimum Y axis dimension of the asset's bounding box\"},\"min_bbox_z\":{\"type\":\"number\",\"title\":\"Min Bbox Z\",\"description\":\"Filter by minimum Z axis dimension of the asset's bounding box\"},\"max_bbox_x\":{\"type\":\"number\",\"title\":\"Max Bbox X\",\"description\":\"Filter by maximum X axis dimension of the asset's bounding box\"},\"max_bbox_y\":{\"type\":\"number\",\"title\":\"Max Bbox Y\",\"description\":\"Filter by maximum Y axis dimension of the asset's bounding box\"},\"max_bbox_z\":{\"type\":\"number\",\"title\":\"Max Bbox Z\",\"description\":\"Filter by maximum Z axis dimension of the asset's bounding box\"},\"return_images\":{\"type\":\"boolean\",\"title\":\"Return Images\",\"description\":\"Return images if set to True\",\"default\":false},\"return_metadata\":{\"type\":\"boolean\",\"title\":\"Return Metadata\",\"description\":\"Return metadata if set to True\",\"default\":false},\"return_root_prims\":{\"type\":\"boolean\",\"title\":\"Return Root Prims\",\"description\":\"Return root prims if set to True\",\"default\":false},\"return_predictions\":{\"type\":\"boolean\",\"title\":\"Return Predictions\",\"description\":\"Return predictions if set to True\",\"default\":false},\"return_in_scene_instances_prims\":{\"type\":\"boolean\",\"title\":\"Return In Scene Instances Prims\",\"description\":\"[in-scene search only] Return prims of instances of objects found in the scene\",\"default\":false},\"embedding_knn_search_method\":{\"allOf\":[{\"$ref\":\"#/components/schemas/SearchMethod\"}],\"description\":\"Search method, approximate should be faster but is less accurate. Default is exact\"},\"limit\":{\"type\":\"integer\",\"maximum\":10000,\"minimum\":1,\"title\":\"Limit\",\"description\":\"Set the maximum number of results to return from the search, default is 32\"},\"vision_metadata\":{\"type\":\"string\",\"title\":\"Vision Metadata\",\"description\":\"Input string for the vision metadata query\"},\"return_vision_generated_metadata\":{\"type\":\"boolean\",\"title\":\"Return Vision Generated Metadata\",\"description\":\"Return vision generated metadata if True\",\"default\":false}},\"type\":\"object\",\"title\":\"DeepSearchSearchRequest\",\"example\":{\"description\":\"box\",\"return_metadata\":true,\"limit\":16}},\"EdgeType\":{\"type\":\"string\",\"enum\":[\"depends_on\",\"parent_prim\",\"root_prim\",\"source_asset\"],\"title\":\"EdgeType\",\"description\":\"An enumeration.\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"minimum\":100,\"maximum\":999,\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"format\":\"^.{1, 1024}$\",\"maxLength\":1024,\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"format\":\"^.{1, 256}$\",\"maxLength\":256,\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36,\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"PaymentRequiredError\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"},\"HealthResponse\":{\"properties\":{\"health_code\":{\"type\":\"string\",\"title\":\"Health Code\",\"description\":\"Status code, OK if successful, otherwise error code.\",\"examples\":[\"200\"]},\"health_code_description\":{\"type\":\"string\",\"title\":\"Health Code Description\",\"description\":\"Status description.\",\"examples\":[\"OK\"]},\"headers\":{\"type\":\"object\",\"title\":\"Headers\",\"description\":\"Response headers.\",\"default\":{\"cache-control\":[\"no-cache\",\"no-store\",\"max-age=0\",\"must-revalidate\"],\"expires\":0,\"pragma\":\"no-cache\"}}},\"type\":\"object\",\"required\":[\"health_code\",\"health_code_description\"],\"title\":\"HealthResponse\"},\"Metadata\":{\"properties\":{\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"}},\"type\":\"object\",\"title\":\"Metadata\"},\"Path\":{\"properties\":{\"version\":{\"type\":\"integer\",\"title\":\"Version\"},\"type\":{\"$ref\":\"#/components/schemas/PathType\"},\"uri\":{\"type\":\"string\",\"title\":\"Uri\"},\"acl\":{\"items\":{\"$ref\":\"#/components/schemas/PathPermission\"},\"type\":\"array\"},\"created\":{\"type\":\"string\",\"title\":\"Created\"},\"created_by\":{\"type\":\"string\",\"title\":\"Created By\"},\"modified\":{\"type\":\"string\",\"title\":\"Modified\"},\"modified_by\":{\"type\":\"string\",\"title\":\"Modified By\"},\"size\":{\"type\":\"number\",\"title\":\"Size\"},\"etag\":{\"type\":\"string\",\"title\":\"Etag\"},\"event\":{\"$ref\":\"#/components/schemas/PathEvent\"},\"mounted\":{\"type\":\"boolean\",\"title\":\"Mounted\"},\"transaction_id\":{\"type\":\"string\",\"title\":\"Transaction Id\"},\"destination\":{\"type\":\"string\",\"title\":\"Destination\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"is_deleted\":{\"type\":\"boolean\",\"title\":\"Is Deleted\"},\"deleted_by\":{\"type\":\"string\",\"title\":\"Deleted By\"},\"deleted_timestamp\":{\"type\":\"number\",\"title\":\"Deleted Timestamp\"}},\"type\":\"object\",\"title\":\"Path\"},\"PathEvent\":{\"enum\":[\"full\",\"create\",\"update\",\"delete\",\"change_acl\",\"set_path_options\",\"lock\",\"unlock\",\"rename\",\"copy\",\"replace_version\"],\"title\":\"PathEvent\",\"description\":\"An enumeration.\"},\"PathPermission\":{\"enum\":[\"read\",\"write\",\"admin\"],\"title\":\"PathPermission\",\"description\":\"An enumeration.\"},\"PathType\":{\"enum\":[\"\",\"asset\",\"folder\",\"channel\",\"mount\",\"object\",\"none\"],\"title\":\"PathType\",\"description\":\"An enumeration.\"},\"Prediction\":{\"properties\":{\"tag\":{\"type\":\"string\",\"title\":\"Tag\"},\"prob\":{\"type\":\"number\",\"title\":\"Prob\"}},\"type\":\"object\",\"required\":[\"tag\",\"prob\"],\"title\":\"Prediction\"},\"Prim\":{\"properties\":{\"scene_url\":{\"type\":\"string\",\"title\":\"Scene Url\"},\"usd_path\":{\"type\":\"string\",\"title\":\"Usd Path\"},\"prim_type\":{\"type\":\"string\",\"title\":\"Prim Type\"},\"bbox_max\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Max\"},\"bbox_min\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Min\"},\"bbox_midpoint\":{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"title\":\"Bbox Midpoint\"},\"bbox_dimension_x\":{\"type\":\"number\",\"title\":\"Bbox Dimension X\"},\"bbox_dimension_y\":{\"type\":\"number\",\"title\":\"Bbox Dimension Y\"},\"bbox_dimension_z\":{\"type\":\"number\",\"title\":\"Bbox Dimension Z\"},\"source_asset_url\":{\"type\":\"string\",\"title\":\"Source Asset Url\"},\"properties\":{\"type\":\"object\",\"title\":\"Properties\"}},\"type\":\"object\",\"required\":[\"scene_url\",\"usd_path\",\"prim_type\",\"bbox_max\",\"bbox_min\",\"bbox_midpoint\",\"bbox_dimension_x\",\"bbox_dimension_y\",\"bbox_dimension_z\"],\"title\":\"Prim\"},\"SearchItem\":{\"properties\":{\"status\":{\"$ref\":\"#/components/schemas/StatusType\"},\"path\":{\"$ref\":\"#/components/schemas/Path\"},\"url\":{\"type\":\"string\",\"title\":\"Url\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"projection\":{\"type\":\"string\",\"title\":\"Projection\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"}},\"type\":\"object\",\"title\":\"SearchItem\"},\"SearchMethod\":{\"type\":\"string\",\"enum\":[\"exact\",\"approximate\"],\"title\":\"SearchMethod\",\"description\":\"An enumeration.\"},\"SearchResult\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"URL of the asset\"},\"score\":{\"type\":\"number\",\"title\":\"Score\"},\"embed\":{\"type\":\"string\",\"title\":\"Embed\"},\"root_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"Root Prims\"},\"image\":{\"type\":\"string\",\"title\":\"Image\"},\"predictions\":{\"items\":{\"$ref\":\"#/components/schemas/Prediction\"},\"type\":\"array\",\"title\":\"Predictions\"},\"vision_generated_metadata\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\",\"title\":\"Vision Generated Metadata\"},\"metadata\":{\"$ref\":\"#/components/schemas/Metadata\"},\"in_scene_instance_prims\":{\"items\":{\"$ref\":\"#/components/schemas/Prim\"},\"type\":\"array\",\"title\":\"In Scene Instance Prims\"}},\"type\":\"object\",\"required\":[\"url\",\"score\"],\"title\":\"SearchResult\",\"examples\":[{\"url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"score\":1.2529583,\"root_prims\":[{\"scene_url\":\"omniverse://simready.ov.nvidia.com/Projects/cardbox_a2.usd\",\"usd_path\":\"/RootNode\",\"prim_type\":\"Xform\",\"bbox_max\":[0.34971755743026733,0.2549635171890259,0.5211517214775085],\"bbox_min\":[-0.34971755743026733,-0.25496378540992737,1.9483268332010084e-8],\"bbox_midpoint\":[0,-1.341104507446289e-7,0.26057587048038844],\"bbox_dimension_x\":0.6994351148605347,\"bbox_dimension_y\":0.5099273025989532,\"bbox_dimension_z\":0.5211517019942402,\"properties\":{\"semantic:QWQQ:params:semanticData\":\"Q1395006\",\"semantic:QWQL:params:semanticType\":\"class\",\"semantic:QWQQ:params:semanticType\":\"qcode\",\"semantic:QWQC:params:semanticData\":\"container/product packaging/box/cardboard box\",\"semantic:QWQL:params:semanticData\":\"cardboard box\",\"semantic:QWQC:params:semanticType\":\"hierarchy\"}}],\"metadata\":{\"created\":\"Mon Mar 20 22:06:58 2023\",\"created_by\":\"user@nvidia.com\",\"modified\":\"Mon Mar 20 22:06:58 2023\",\"modified_by\":\"user@nvidia.com\",\"size\":14938,\"etag\":\"169176\"},\"vision_generated_metadata\":{\"vision_generated_object_type\":\"electric guitar, musical instrument, guitar\",\"vision_generated_materials\":\"wood, metal, plastic\"}}]},\"StatusType\":{\"enum\":[\"OK\",\"DENIED\",\"TOKEN_EXPIRED\",\"UNAUTHORIZED\",\"ES_REQUEST_ERROR\",\"ES_CONNECTION_TIMEOUT\",\"FILE_NOT_FOUND_ERROR\",\"THUMBNAIL_MISSING_ERROR\",\"INVALID_PREFIX\",\"UNKNOWN_ERROR\",\"PROJECTION_SERVICE_UNAVAILABLE\",\"REST_API_UNAVAILABLE\"],\"title\":\"StatusType\",\"description\":\"An enumeration.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}},\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T16:39:16.212Z\",\"nvcfFunctionId\":\"f2edbb70-232f-4a38-89c1-6eb5dbf92eed\",\"createdDate\":\"2024-11-04T17:12:32.654Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-usdsearch\",\"blueprintUrl\":\"https://github.com/NVIDIA-Omniverse-blueprints/3d-conditioning\",\"termsOfUse\":\"GOVERNING TERMS: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). ADDITIONAL INFORMATION: RealvisXL license at [LICENSE.md · stabilityai/stable-diffusion-xl-base-1.0 at main.](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/blob/main/LICENSE.md).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Run Anywhere - Notify Me\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"projects\":[{\"name\":\"Get started with this NVIDIA NIM microservice today\",\"url\":\"https://docs.omniverse.nvidia.com/usdsearch/get-started.html\",\"imageUrl\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/usdsearch.jpg\",\"workbench\":false}]},\"artifactName\":\"conditioning-for-precise-visual-generative-ai\"},\"config\":{\"name\":\"conditioning-for-precise-visual-generative-ai\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"displayName\":\"Synthetic Manipulation Motion Generation for Robotics\",\"publisher\":\"nvidia\",\"shortDescription\":\"Generate exponentially large amounts of synthetic motion trajectories for robot manipulation from just a few human demonstrations.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/isaac-gr00t-synthetic-manipulation.jpg\",\"labels\":[\"Blueprint\",\"Humanoids\",\"NVIDIA Isaac GR00T\",\"image-to-world\",\"physical ai\",\"robot learning\",\"robotics\",\"synthetic data\",\"teleop\",\"text-to-world\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"true\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:11:29.060Z\",\"description\":\"$59\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-19T18:02:34.533Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"93003d0b-a573-4906-99ef-82a0f14cf24d\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Example - AI Virtual Assistant for Customer Service\",\"description\":\"This API schema describes all the endpoints exposed by the AI Virtual Assistant for Customer Service NIM Blueprint\",\"version\":\"1.0.0\"},\"paths\":{\"/agent/metrics\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Get Metrics\",\"operationId\":\"get_metrics_agent_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/agent/health\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Health Check\",\"description\":\"Perform a Health Check\\n\\nReturns 200 when service is up. This does not check the health of downstream services.\",\"operationId\":\"health_check_agent_health_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthResponse\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}}}},\"/agent/generate\":{\"post\":{\"tags\":[\"Agent\"],\"summary\":\"Generate Response\",\"description\":\"Generate and stream the response to the provided prompt.\",\"operationId\":\"generate_response_agent_generate_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}},\"x-nvai-meta\":{\"examples\":[{\"name\":\"What was my last purchase?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What was my last purchase?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"},{\"name\":\"What product did I order first?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What product did I order first?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"}]}}}},\"components\":{\"schemas\":{\"AgentRequest\":{\"properties\":{\"messages\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":50000},{\"type\":\"null\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between user and assistant. The last input message should have role user. A message with the the system role is optional, and must be the very first message if it is present.\",\"default\":[]},\"user_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"User Id\",\"description\":\"A unique identifier representing your end-user.\",\"default\":\"\"},\"session_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\",\"default\":\"\"},\"api_type\":{\"type\":\"string\",\"title\":\"Api Type\",\"description\":\"The type of API action: 'create_session', 'end_session', 'generate', or 'summary'.\"}},\"type\":\"object\",\"required\":[\"api_type\"],\"title\":\"AgentRequest\",\"description\":\"Definition of the Prompt API data type.\"},\"AgentResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"maxLength\":100000,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Id\",\"default\":\"\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/AgentResponseChoices\"},\"type\":\"array\",\"maxItems\":256,\"title\":\"Choices\",\"default\":[]},\"session_id\":{\"type\":\"string\",\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\"}},\"type\":\"object\",\"title\":\"AgentResponse\",\"description\":\"Definition of Chain APIs resopnse data type\"},\"AgentResponseChoices\":{\"properties\":{\"index\":{\"type\":\"integer\",\"maximum\":256,\"minimum\":0,\"format\":\"int64\",\"title\":\"Index\",\"default\":0},\"message\":{\"$ref\":\"#/components/schemas/Message\",\"default\":{\"role\":\"assistant\",\"content\":\"\"}},\"finish_reason\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Finish Reason\",\"default\":\"\"},\"sentiment\":{\"type\":\"string\",\"title\":\"Sentiment\",\"description\":\"Any sentiment associated with this message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"AgentResponseChoices\",\"description\":\"Definition of Chain response choices\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthResponse\":{\"properties\":{\"message\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"HealthResponse\"},\"Message\":{\"properties\":{\"role\":{\"type\":\"string\",\"maxLength\":256,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Role\",\"description\":\"Role for a message AI, User and System\",\"default\":\"user\"},\"content\":{\"type\":\"string\",\"maxLength\":131072,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Content\",\"description\":\"The input query/prompt to the pipeline.\",\"default\":\"I am going to Paris, what should I see?\"}},\"type\":\"object\",\"title\":\"Message\",\"description\":\"Definition of the Chat Message type.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}},\"tags\":[{\"name\":\"Health\",\"description\":\"APIs for checking and monitoring server liveliness and readiness.\"},{\"name\":\"Agent\",\"description\":\"Core APIs for interacting with the agent.\"}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-19T18:02:35.234Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:11:29.392Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; use of the model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License Agreement\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-Omniverse-blueprints/synthetic-motion-generation\"}},\"artifactName\":\"isaac-gr00t-synthetic-manipulation\"},\"config\":{\"name\":\"isaac-gr00t-synthetic-manipulation\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"displayName\":\"Test Multi-Robot Fleets for Industrial Automation\",\"publisher\":\"nvidia\",\"shortDescription\":\"Simulate, test, and optimize physical AI and robotic fleets at scale in industrial digital twins before real-world deployment.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/mega-multi-robot-fleets-for-industrial-automation.jpg\",\"labels\":[\"Blueprint\",\"NVIDIA Omniverse\",\"industrial\",\"omniverse blueprint\",\"simulation\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"true\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:09:17.934Z\",\"description\":\"$5a\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-31T16:01:59.380Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"efdb7052-704c-49eb-bd69-a854d79ecc21\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Example - AI Virtual Assistant for Customer Service\",\"description\":\"This API schema describes all the endpoints exposed by the AI Virtual Assistant for Customer Service NIM Blueprint\",\"version\":\"1.0.0\"},\"paths\":{\"/agent/metrics\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Get Metrics\",\"operationId\":\"get_metrics_agent_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/agent/health\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Health Check\",\"description\":\"Perform a Health Check\\n\\nReturns 200 when service is up. This does not check the health of downstream services.\",\"operationId\":\"health_check_agent_health_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthResponse\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}}}},\"/agent/generate\":{\"post\":{\"tags\":[\"Agent\"],\"summary\":\"Generate Response\",\"description\":\"Generate and stream the response to the provided prompt.\",\"operationId\":\"generate_response_agent_generate_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}},\"x-nvai-meta\":{\"examples\":[{\"name\":\"What was my last purchase?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What was my last purchase?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"},{\"name\":\"What product did I order first?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What product did I order first?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"}]}}}},\"components\":{\"schemas\":{\"AgentRequest\":{\"properties\":{\"messages\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":50000},{\"type\":\"null\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between user and assistant. The last input message should have role user. A message with the the system role is optional, and must be the very first message if it is present.\",\"default\":[]},\"user_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"User Id\",\"description\":\"A unique identifier representing your end-user.\",\"default\":\"\"},\"session_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\",\"default\":\"\"},\"api_type\":{\"type\":\"string\",\"title\":\"Api Type\",\"description\":\"The type of API action: 'create_session', 'end_session', 'generate', or 'summary'.\"}},\"type\":\"object\",\"required\":[\"api_type\"],\"title\":\"AgentRequest\",\"description\":\"Definition of the Prompt API data type.\"},\"AgentResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"maxLength\":100000,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Id\",\"default\":\"\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/AgentResponseChoices\"},\"type\":\"array\",\"maxItems\":256,\"title\":\"Choices\",\"default\":[]},\"session_id\":{\"type\":\"string\",\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\"}},\"type\":\"object\",\"title\":\"AgentResponse\",\"description\":\"Definition of Chain APIs resopnse data type\"},\"AgentResponseChoices\":{\"properties\":{\"index\":{\"type\":\"integer\",\"maximum\":256,\"minimum\":0,\"format\":\"int64\",\"title\":\"Index\",\"default\":0},\"message\":{\"$ref\":\"#/components/schemas/Message\",\"default\":{\"role\":\"assistant\",\"content\":\"\"}},\"finish_reason\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Finish Reason\",\"default\":\"\"},\"sentiment\":{\"type\":\"string\",\"title\":\"Sentiment\",\"description\":\"Any sentiment associated with this message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"AgentResponseChoices\",\"description\":\"Definition of Chain response choices\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthResponse\":{\"properties\":{\"message\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"HealthResponse\"},\"Message\":{\"properties\":{\"role\":{\"type\":\"string\",\"maxLength\":256,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Role\",\"description\":\"Role for a message AI, User and System\",\"default\":\"user\"},\"content\":{\"type\":\"string\",\"maxLength\":131072,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Content\",\"description\":\"The input query/prompt to the pipeline.\",\"default\":\"I am going to Paris, what should I see?\"}},\"type\":\"object\",\"title\":\"Message\",\"description\":\"Definition of the Chat Message type.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}},\"tags\":[{\"name\":\"Health\",\"description\":\"APIs for checking and monitoring server liveliness and readiness.\"},{\"name\":\"Agent\",\"description\":\"Core APIs for interacting with the agent.\"}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-31T16:01:59.959Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:09:18.304Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; use of the model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License Agreement\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Notify When Available\",\"url\":\"https://www.nvidia.com/en-us/omniverse/mega-multi-robot-fleets-notify-me/\"}},\"artifactName\":\"mega-multi-robot-fleets-for-industrial-automation\"},\"config\":{\"name\":\"mega-multi-robot-fleets-for-industrial-automation\",\"type\":\"blueprint\"}}],\"items\":[\"$5b\",\"$5c\",\"$5d\",\"$5e\",\"$5f\",\"$60\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"Accelerate Your Simulation Workflows\"}],[\"$\",\"$L26\",null,{\"href\":\"/blueprints?filters=blueprintType%3Ablueprinttype_nvidia_omniverse\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"Blueprints to help you expedite simulation and development with NVIDIA Omniverse.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"61:T12e7,"])</script><script>self.__next_f.push([1,"# Genomics Analysis\nEasily GPU accelerate essential genomics analysis workflows, such as germline, by using NVIDIA Parabricks.\n\n## Overview\n\nFor germline analysis, bioinformaticians can try a whole exome sequencing analysis workflow on short reads in a matter of minutes on any cloud available through Brev.dev, leveraging [NVIDIA® Parabricks®](https://docs.nvidia.com/clara/parabricks/latest/index.html) fq2bam (BWA-MEM) for alignment and DeepVariant for variant calling. \n\n## Experience Workflow\n\nIt is strongly recommended that users review the [README](https://github.com/clara-parabricks-workflows/genomics-analysis-blueprint/blob/main/README.md) in this blueprint before working through the notebooks. Users can then execute the experience workflow in the germline_wes notebook.\n\nThe workflow is as follows:\n1. This example uses whole exome sequencing (WES) data from sample NA12878. \n2. Sequence reads are mapped to the reference genome. The input FASTQ files are aligned using the Burrows-Wheeler Aligner (BWA) through the Parabricks fq2bam tool. \n3. Run DeepVariant, a deep learning-based variant caller, on the aligned reads. It uses a convolutional neural network to find single nucleotide variants (SNVs) and insertions/deletions (InDels). \n\n## Architecture Diagram\n\nShort-Read Analysis Workflow\n\n![](https://assets.ngc.nvidia.com/products/api-catalog/genomics-analysis/diagram.jpg)\n\n## Software\n\n* The Parabricks 4.4.0 container is used in this blueprint - [try the latest container here](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/containers/clara-parabricks).\n\n## Reference(s)\n\n* [Technical Documentation](https://docs.nvidia.com/clara/parabricks/4.4.0/index.html) \n* [Parabricks 4.4.0 Technical Blog Post (including benchmarks)](https://developer.nvidia.com/blog/discover-new-biological-insights-with-accelerated-pangenome-alignment-in-nvidia-parabricks/)\n\n## Minimum System Requirements\n\nUsers may have to wait 5–10 minutes for the instance to start, depending on cloud availability. The germline analysis blueprint supports the following hardware: \n\n**Hardware Requirements**\n\n* The L40s with at least 48GB of GPU memory is recommended for the best combination of cost and performance. Users can also try L4 or T4 (better cost) or A100 (better performance). \n* The [fq2bam](https://docs.nvidia.com/clara/parabricks/latest/Documentation/ToolDocs/man_fq2bam.html#man-fq2bam) tool requires at least 40 GB of GPU memory by default; the `--low-memory` option will reduce this to 16GB of GPU memory at the cost of slower processing. All other tools require at least 16GB of GPU memory per GPU.\n\n**Optional Hardware Requirements**\n\n* Parabricks can be run on any NVIDIA GPU that supports CUDA® architecture 70, 75, 80, 86, 89, or 90 and has at least 16GB of GPU RAM. NVIDIA Parabricks has been tested on the following NVIDIA GPUs: \n * V100 \n * T4 \n * A10, A30, A40, A100, A6000 \n * L4, L40 \n * H100, H200 \n * Grace Hopper™ Superchip \n* System Requirements: \n * A 2-GPU system should have at least 100GB of CPU RAM and at least 24 CPU threads. \n * A 4-GPU system should have at least 196GB of CPU RAM and at least 32 CPU threads. \n * An 8-GPU system should have at least 392GB of CPU RAM and at least 48 CPU threads.\n\n**Software Requirements**\n\n* An NVIDIA driver with version 525.60.13 or greater. Please check [here](https://docs.nvidia.com/deploy/cuda-compatibility/#forward-compatibility) for more details on forward compatibility. \n* Any Linux operating system that supports Docker version 20.10 (or higher) with the NVIDIA GPU runtime.\n\n## Terms of Use\n\n**Governing Terms**: The Parabricks container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product-Specific Terms for NVIDIA AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/). This Genomics Analysis Blueprint github repository is provided under [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0).\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"62:T1aee,"])</script><script>self.__next_f.push([1,"# Single-Cell Analysis\n\nInvestigate, understand, and interpret single-cell data in minutes, not days, by leveraging RAPIDS-singlecell, powered by NVIDIA RAPIDS™\n\n## Overview\n\nFor single-cell analysis, scientists can test near-real-time data analysis and visualization easily, achieving up to 938X faster accelerations versus CPU by using [RAPIDS-singlecell](https://rapids-singlecell.readthedocs.io/), developed by [scverse](https://scverse.org/about/). This blueprint is for scientists who understand single-cell analysis and want to leverage [RAPIDS](https://rapids.ai/) for single-cell data.\n\n## Experience Workflow\n\nIt is strongly recommended that users review the [README](https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint/blob/main/README.md) in this blueprint before working through the notebooks.\n\nFor this blueprint, two possible deployments are provided:\n\n1. The Standard Instance: L40s \n2. The Large Instance: 8x H100\n\nPlease use the table in the Notebook Overview below to determine which size is right for you.\n\nThe workflow is as follows:\n\n1. After initial code setup, this blueprint utilizes publicly available datasets including those from [10x Genomics](https://www.10xgenomics.com/datasets) and [CZ CELLxGENE](https://cellxgene.cziscience.com/). Scientists can use their [Python API](https://chanzuckerberg.github.io/cellxgene-census/python-api.html#) to read the data directly into an [AnnData](https://anndata.readthedocs.io/en/stable/) object. \n2. General data preprocessing is performed to clean up and better understand the dataset. This includes calculating QC metrics, filtering, and data normalization. \n3. The data is investigated quantitatively and visually, including feature selection, clustering, dimensionality reduction, and data integration using canonical tools. \n4. The data is visualized and plotted to help users investigate the biological diversity within the sample. \n5. A number of additional advanced tutorials are available for users who are interested in spatial transcriptomics analysis, as well as scaling to 11M cells easily and quickly.\n\n### **Notebooks Outline**\n\nThe outline below is a suggested exploration flow. Unless otherwise noted, users can choose any notebook to get started, as long as the GPU resources are available to run the notebook.\n\nFor those who are new to doing basic analysis for single-cell data, the end-to-end analysis of [01\\_demo\\_gpu\\_e2e](https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint/blob/main/01_demo_gpu_e2e.ipynb) is the best place to start, where users are walked through the steps of data preprocessing, cleanup, visualization, and investigation.\n\n| Notebook | Description | Min GPU Size / Instance |\n| :---- | :---- | :---- |\n| 01\\_demo\\_gpu\\_e2e | End-to-end workflow, where we understand the cells, run ETL on the dataset then visualize and explore the results. This tutorial is good for all users. | 24GB / Standard Instance |\n| 02\\_decoupler | This notebook continues from the outputs of 01\\_demo\\_gpu\\_e2e as an overview of methods that can be used to investigate transcriptional regulation. | 24GB / Standard Instance |\n| demo\\_gpu\\_e2e\\_with\\_PR | End-to-end workflow, like 01\\_demo\\_gpu\\_e2e, but uses Pearson residuals for normalization. | 24GB / Standard Instance |\n| spatial\\_autocorr | An introduction to spatial transcriptomics analysis and visualization. | 24GB / Standard Instance |\n| out-of-core\\_processing | In this notebook, we show the scalability of the analysis of up to 11M cells easily by using Dask. Requires a 48GB GPU. | 48GB / Standard Instance |\n| multi\\_gpu\\_large\\_data\\_showcase | This notebook enhances the 11M cell dataset analysis with Dask without exceeding memory limits. It fully scales to utilize all available GPUs, uses chunk-based execution, and efficiently manages memory. Requires 8x H100s or better. For all other GPU systems, please run out-of-core\\_processing instead. | 8x 80GB / Large Instance |\n| demo\\_gpu-seuratv3 | In this notebook, show diversity in capability by running a similar workflow to 01\\_demo\\_gpu\\_e2e but on brain cells. | 24GB / Standard Instance |\n| demo\\_gpu-seuratv3-brain-1M | In this notebook, we scale up the analysis of demo\\_gpu-seuratv3 to 1 million brain cells. Requires an 80GB GPU, like an H100. | 80GB / Large Instance |\n\n## Architecture Diagram\n\n## ![](https://assets.ngc.nvidia.com/products/api-catalog/single-cell-analysis/diagram.jpg)\n\n## Software\n\nThe following containers are used in this blueprint:\n\n* [RAPIDS](https://developer.nvidia.com/rapids) v24.12\n\nAdditional software—including use of [RAPIDS-singlecell](https://rapids-singlecell.readthedocs.io/en/latest/), developed by [scverse](https://scverse.org/about/)—[is available on GitHub accompanying these notebooks](https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint/).\n\n## Minimum System Requirements\n\nThe single-cell analysis blueprint recommends using L40s with minimum 24GB VRAM, unless otherwise stated in the tutorial. Users may have to wait 5–10 minutes for the instance to start, depending on cloud availability.\n\nThe blueprint supports:\n\nHardware Requirements\n\n* We recommend using NVIDIA GPU L40s for the best user experience and performance-to-cost ratio for this blueprint, unless otherwise stated in the tutorial. The Large or MultiGPU notebooks require one or more 80GB GPUs. We suggest using an 8x H100 instance. \n* Other supported instances, if available in your region: \n * H100 \n * A100 \n * A10 \n * L4 \n * GH200 \n* 24 GB VRAM or more recommended\n\nSoftware Requirements\n\n* [Environment packages can be found on GitHub](https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint/)\n\n## License\n\nGoverning Terms:\n\n* This RAPIDS-singlecell Blueprint GitHub repository is provided under [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0). \n* RAPIDS projects are released under the [Apache-2.0 license](https://docs.rapids.ai/contributing/code/#new-developers). \n* The RAPIDS-singlecell license is [available here](https://github.com/scverse/rapids_singlecell/blob/main/LICENSE).\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"63:T156c,"])</script><script>self.__next_f.push([1,"## Use Case Description\n\nIn drug discovery, even if you know which protein to target that would treat a disease, designing a therapeutic molecule that specifically binds that protein is a staggering challenge.\nImagine searching for a single, perfectly shaped key in a warehouse of nearly infinite keys—each with a unique three-dimensional shape. This isn't just a metaphor; for a protein of length ‘n’, there can be 20^n possible sequences, each capable of adopting countless conformations. Since the average human protein is 430 amino acids, this represents 20^430 possible sequences, a practically infinite number and more than the number of atoms in the universe (10^80). This potential diversity is so important in the evolution of life but presents a challenge for researchers. \nIn traditional workflows, this complexity means painstaking trial and error—iterating through thousands of candidates, each synthesis and validation round taking months, if not years. The process is expensive, slow, and fraught with uncertainty. Researchers often use educated guesses and hope that a binder emerges from the colossal search.\nHere, we bring generative AI to bear on the problem, pre-optimizing molecules and screening their interaction with the target protein. This BioNeMo blueprint shows how protein binder design can be recast using NIM microservices for protein folding, structure generation, and sequence generation to speed up the development cycle and produce better binders faster.\n\n## Experience Walkthrough\n\n\n1. The Protein Binder Design NVIDIA NIM Agent Blueprint leverages AI models packaged within NIM microservices to design optimized protein sequences and structures. The workflow begins with the user providing an amino acid sequence to AlphaFold2, which predicts the initial 3D structure of the target protein. AlphaFold2 also requires a multi-sequence alignment, which can be generated with an accelerated MSA NIM. \n2. The structure of the protein target is then used by RFdiffusion to design a protein binder. At this stage, RFdiffusion generates only the backbone of the protein binder. The model can be steered by the user to explore specific binding interfaces, or hot spot regions of the target protein and identify the most favorable binding configurations according to the user’s desired design constraints. \n3. Next, ProteinMPNN generates and optimizes amino acid sequences that fit into the RFdiffusion-generated protein backbone, ensuring they exhibit the necessary biochemical properties for effective binding. \n4. Finally, AlphaFold2-Multimer is used to validate the interactions and stability of the resulting protein complexes. This integrated approach enables the precise and efficient design of protein binders, facilitating advancements in therapeutic protein development and other protein engineering applications.\n\n\n\n## What's included in the Blueprint\n \nNVIDIA AI Blueprints are customizable AI workflow examples that equip enterprise developers with NIM microservices, reference code, documentation, and a Helm chart for deployment. \n\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/proteindesign/diagram.png)\n\n## Minimum System Requirements\n\n### Hardware\n\n- GPU: 4 or more NVIDIA L40s, A100, or H100 GPUs, each with at least 48 GB of VRAM\n- CPU: x86_64 architecture only for this release, with 24 physical CPU cores\n- Storage: 1300GB of fast NVMe SSD storage \n- System Memory: 64GB\n\n### Software\n\n- Operating System: Ubuntu 20.04 or newer\n- NVIDIA Driver version: 535 or newer\n- NVIDIA CUDA® version: 12.4 or newer\n- NVIDIA Container Toolkit version: 1.15.0 or newer\n- Docker version: Docker version 26 or newer\n- Python Version 3.11+\n\n\n## Example Walkthrough with Sample Input/Output\n\nSee a complete example of how to get started with this blueprint on the [NVIDIA BioNeMo Blueprints GitHub repository](https://github.com/NVIDIA-BioNeMo-blueprints/generative-protein-binder-design)\n\n## Technical Considerations\n\nRFdiffusion is capable of generating protein backbones, and ProteinMPNN can label the amino acid sequence. The combination yields sequences that should fold into the protein structure that binds to the specified static target protein structure. Note that proteins are flexible and adopt multiple conformations. This is especially true of antibodies where the binding interface is disordered. This poses a challenge for these models. While AI is making great strides in predicting protein sequences and structures that should bind to target proteins, it’s not perfect.\n\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## License\n\nUse of the models in this Blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf)."])</script><script>self.__next_f.push([1,"64:T14fd,"])</script><script>self.__next_f.push([1,"## Use Case Description\nComputational drug designers must pick a few chemical structures from around 10^60 options for experimental testing, more than the number of stars in the universe. To discover \"hits\" that have all the properties of a drug suitable for clinical testing, their search must be targeted and efficient.\n\nThis is a difficult problem, and pharma companies typically spend 10-15 years and $1B-$2B to bring a new drug to solve it and bring a new drug to market. In a typical drug discovery workflow, researchers first identify the biological target and mechanism that they want to alter to treat the disease, a process called target identification. Then, once a target is identified, molecules that bind to that target must be identified (hit identification). \nThese hits are then optimized for safety and therapeutic effect.\n\nThe biology and chemistry underlying each of these steps is complex, often involving the identification of cryptic patterns in enormous datasets and long cycles of biological experimentation, chemical synthesis and validation. However, even though Pharma spent $262B USD on R\u0026D in 2023 (Evaluate), 90% of drugs in clinical trials fail, demonstrating a need for innovative approaches to drug discovery (Nature Review Drug Discovery).\n\nHere, we bring generative AI to bear on the problem, pre-optimizing molecules and screening their interaction with the target protein. This NIM Agent Blueprint shows how virtual screening can be recast using NVIDIA microservices for protein folding, molecule generation, and docking to speed the development cycle and produce better molecules, faster.\n\n\n## Experience Walkthrough\n\n\n1. The user passes the sequence of the protein target that they want to design against to the OpenFold2 NIM, which accurately determines that protein's structure. This step uses an alignment of the protein sequence to other known proteins with the MMseqs2 NIM, and multiple configurations for this alignment step are available.\n\n2. An initial library of chemical fragments is passed to the GenMol NIM to seed its generative chemical design. The user can also choose a property to optimize for (e.g., QED), the number of molecules to generate, and other constraints. The generated molecules are scored and passed back to GenMol for further optimization for multiple cycles depending on the number of iterations the user selects.\n\n3. These molecular structures and the structure of the protein target are passed to the DiffDock NIM, which generates the number of binding poses that the user indicates, along with other constraints.\n\n4. The user then clicks the \"Generate Molecules\" button, and when complete, optimized molecules are returned to the user, ready for further lab testing.\n\n## What's included in the Blueprint\n\nNVIDIA NIM\u003csup\u003eTM\u003c/sup\u003e Agent Blueprints are customizable AI workflow examples that equip enterprise developers with NIM microservices, reference code, documentation, and a Helm chart for deployment. \n\n![Architecture Diagram Placeholder](https://assets.ngc.nvidia.com/products/api-catalog/generative-virtual-screening-for-drug-discovery/caddblueprint.png)\n\n## Minimum System Requirements\n\n### Hardware\n- At least 1300 GB (1.3 TB) of fast NVMe SSD space. (For MSA databases)\n- A modern CPU with at least 24 CPU cores\n- At least 64 GB of RAM\n- 4 X NVIDIA L40s, A100, or H100 GPUs across your cluster.\n\n### Software\n\n- Operating System: Ubuntu 20.04 or newer \n- NVIDIA Driver version: 535 or newer \n- NVIDIA CUDA version: 12.4 or newer \n- NVIDIA Container Toolkit version: 1.15.0 or newer \n- Docker version: Docker version 26 or newer\n\n## Example Walkthrough with Sample Input/Output\n\nSee a complete example of how to get started with this blueprint on the [NVIDIA BioNeMo Blueprints GitHub repository](https://github.com/NVIDIA-NIM-Agent-Blueprints/generative-virtual-screening)\n\n## Technical Considerations\n\nGenMol can optimize molecules for user-defined objectives by integrating experimental or computational feedback via oracles. Using discrete diffusion and fragment remasking, it iteratively modifies molecular fragments while preserving key structures to enhance properties like binding affinity, QED, and SA. To use GenMol for optimization, deploy the GenMol NIM and utilize the ‘/generate’ endpoint to guide molecular design with custom oracle functions. Learn more in the GenMol NIM documentation and example Jupyter notebooks.\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## License\n\nUse of the models in this Generative Virtual Screening Blueprint are governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf)."])</script><script>self.__next_f.push([1,"65:{\"name\":\"genomics-analysis\",\"type\":\"blueprint\"}\n66:{\"name\":\"single-cell-analysis\",\"type\":\"blueprint\"}\n67:{\"name\":\"protein-binder-design-for-drug-discovery\",\"type\":\"blueprint\"}\n68:{\"name\":\"generative-virtual-screening-for-drug-discovery\",\"type\":\"blueprint\"}\n"])</script><script>self.__next_f.push([1,"3c:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"genomics-analysis\",\"displayName\":\"Genomics Analysis\",\"publisher\":\"nvidia\",\"shortDescription\":\"Easily run essential genomics workflows to save time leveraging Parabricks\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/genomics-analysis.jpg\",\"labels\":[\"Biology\",\"Blueprint\",\"Genomics\",\"Parabricks\",\"NVIDIA AI\",\"DNA Sequencing\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:18:39.575Z\",\"description\":\"$61\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:19:09.891Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"2247a25e-c2a3-4bbc-bd9b-3daf5cdb304a\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:19:10.314Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:18:39.788Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"By using this software or model, you are agreeing to the NVIDIA Parabricks \u003ca href=\\\"https://docs.nvidia.com/clara/parabricks/latest/documentation/eula.html\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eTerms of Use.\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://nvda.ws/41BxA49\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/clara-parabricks-workflows/genomics-analysis-blueprint\"},\"notebook\":\"\"},\"artifactName\":\"genomics-analysis\"},\"config\":{\"name\":\"genomics-analysis\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"single-cell-analysis\",\"displayName\":\"Single Cell Analysis\",\"publisher\":\"nvidia\",\"shortDescription\":\"Investigate, understand, and interpret single cell data in minutes, not days by leveraging RAPIDS-singlecell, powered by NVIDIA RAPIDS\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/single-cell-analysis.jpg\",\"labels\":[\"Biology\",\"Blueprint\",\"Genomics\",\"RAPIDS\",\"Single Cell\",\"NVIDIA AI\",\"RNA Sequencing\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:18:41.993Z\",\"description\":\"$62\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:18:41.993Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"eda87d46-8b4f-4f0c-9a2f-495d2299de27\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Example - AI Virtual Assistant for Customer Service\",\"description\":\"This API schema describes all the endpoints exposed by the AI Virtual Assistant for Customer Service NIM Blueprint\",\"version\":\"1.0.0\"},\"paths\":{\"/agent/metrics\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Get Metrics\",\"operationId\":\"get_metrics_agent_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/agent/health\":{\"get\":{\"tags\":[\"Health\"],\"summary\":\"Health Check\",\"description\":\"Perform a Health Check\\n\\nReturns 200 when service is up. This does not check the health of downstream services.\",\"operationId\":\"health_check_agent_health_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthResponse\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}}}},\"/agent/generate\":{\"post\":{\"tags\":[\"Agent\"],\"summary\":\"Generate Response\",\"description\":\"Generate and stream the response to the provided prompt.\",\"operationId\":\"generate_response_agent_generate_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/AgentResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}},\"500\":{\"description\":\"Internal Server Error\",\"content\":{\"application/json\":{\"example\":{\"detail\":\"Internal server error occurred\"}}}}},\"x-nvai-meta\":{\"examples\":[{\"name\":\"What was my last purchase?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What was my last purchase?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"},{\"name\":\"What product did I order first?\",\"requestJson\":\"{\\n \\\"user\\\": \\\"1\\\", \\\"api_type\\\": \\\"create_session\\\", \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What product did I order first?\\\"\\n }\\n ],\\n \\\"stream\\\": true\\n}\\n\"}]}}}},\"components\":{\"schemas\":{\"AgentRequest\":{\"properties\":{\"messages\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":50000},{\"type\":\"null\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between user and assistant. The last input message should have role user. A message with the the system role is optional, and must be the very first message if it is present.\",\"default\":[]},\"user_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"User Id\",\"description\":\"A unique identifier representing your end-user.\",\"default\":\"\"},\"session_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\",\"default\":\"\"},\"api_type\":{\"type\":\"string\",\"title\":\"Api Type\",\"description\":\"The type of API action: 'create_session', 'end_session', 'generate', or 'summary'.\"}},\"type\":\"object\",\"required\":[\"api_type\"],\"title\":\"AgentRequest\",\"description\":\"Definition of the Prompt API data type.\"},\"AgentResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"maxLength\":100000,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Id\",\"default\":\"\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/AgentResponseChoices\"},\"type\":\"array\",\"maxItems\":256,\"title\":\"Choices\",\"default\":[]},\"session_id\":{\"type\":\"string\",\"title\":\"Session Id\",\"description\":\"A unique identifier representing the session associated with the response.\"}},\"type\":\"object\",\"title\":\"AgentResponse\",\"description\":\"Definition of Chain APIs resopnse data type\"},\"AgentResponseChoices\":{\"properties\":{\"index\":{\"type\":\"integer\",\"maximum\":256,\"minimum\":0,\"format\":\"int64\",\"title\":\"Index\",\"default\":0},\"message\":{\"$ref\":\"#/components/schemas/Message\",\"default\":{\"role\":\"assistant\",\"content\":\"\"}},\"finish_reason\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Finish Reason\",\"default\":\"\"},\"sentiment\":{\"type\":\"string\",\"title\":\"Sentiment\",\"description\":\"Any sentiment associated with this message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"AgentResponseChoices\",\"description\":\"Definition of Chain response choices\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthResponse\":{\"properties\":{\"message\":{\"type\":\"string\",\"maxLength\":4096,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Message\",\"default\":\"\"}},\"type\":\"object\",\"title\":\"HealthResponse\"},\"Message\":{\"properties\":{\"role\":{\"type\":\"string\",\"maxLength\":256,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Role\",\"description\":\"Role for a message AI, User and System\",\"default\":\"user\"},\"content\":{\"type\":\"string\",\"maxLength\":131072,\"pattern\":\"[\\\\s\\\\S]*\",\"title\":\"Content\",\"description\":\"The input query/prompt to the pipeline.\",\"default\":\"I am going to Paris, what should I see?\"}},\"type\":\"object\",\"title\":\"Message\",\"description\":\"Definition of the Chat Message type.\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}},\"tags\":[{\"name\":\"Health\",\"description\":\"APIs for checking and monitoring server liveliness and readiness.\"},{\"name\":\"Agent\",\"description\":\"Core APIs for interacting with the agent.\"}]},\"namespace\":\"qc69jvmznzxy\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:18:42.305Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"License is available on GitHub \u003ca href=\\\"https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e \\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy\",\"menu\":[{\"text\":\"Deploy Standard Instance\",\"url\":\"https://nvda.ws/3DzqANc\"},{\"text\":\"Deploy Large Instance\",\"url\":\"https://nvda.ws/3DwSxoU\"}]},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/clara-parabricks-workflows/single-cell-analysis-blueprint\"},\"notebook\":\"\"},\"artifactName\":\"single-cell-analysis\"},\"config\":{\"name\":\"single-cell-analysis\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"protein-binder-design-for-drug-discovery\",\"displayName\":\"Build A Generative Protein Binder Design Pipeline\",\"publisher\":\"nvidia\",\"shortDescription\":\"This blueprint shows how generative AI and accelerated NIM microservices can design protein binders smarter and faster.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/protein-binder-design-for-drug-discovery.jpg\",\"labels\":[\"BioNemo\",\"Biology\",\"Blueprint\",\"Protein Generation\",\"NVIDIA BioNemo\",\"Drug Discovery\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"NIM\",\"value\":\"rfdiffusion,proteinmpnn,alphafold2,alphafold2-multimer\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-13T17:26:05.381Z\",\"description\":\"$63\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T17:40:05.360Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"535f0546-86b9-481f-8af0-b5e525e677f4\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Protein Design\",\"version\":\"0.1.0\"},\"paths\":{\"/protein-design/design-binders-from-sequence\":{\"post\":{\"summary\":\"Design Binders\",\"operationId\":\"design_binders_protein_design_design_binders_from_sequence_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ProteinDesignGenerationInputs\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ProteinDesignGenerationOutputs\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}}}},\"/v2/health/ready\":{\"get\":{\"summary\":\"Health Check\",\"operationId\":\"health_check_v2_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"ProteinDesignGenerationInputs\":{\"properties\":{\"sequence\":{\"type\":\"string\",\"maxLength\":4096,\"minLength\":1,\"pattern\":\"^[ARNDCQEGHILKMFPSTWYV]+\",\"title\":\"Input Polypeptide Sequence\",\"description\":\"An input polypeptide (i.e., amino acid) sequence that must be composed of valid Amino Acid IUPAC symbols.\"},\"msa_algorithm\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"jackhmmer\",\"mmseqs2\"]},{\"type\":\"null\"}],\"title\":\"MSA Algorithm\",\"description\":\"The algorithm to use for MSA. AlphaFold2 was trained on JackHMMer; MMSeqs2 provides faster inference.\",\"default\":\"jackhmmer\",\"tooltip\":\"MSA Algorithm used for generating model MSA features.\"},\"e_value\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}],\"title\":\"MSA e-value.\",\"description\":\"The e-value used for filtering hits when building the Multiple Sequence Alignment.\",\"default\":0.0001,\"tooltip\":\"Sequence e-value for filtering sequences seeding the MSA.\"},\"relax_prediction\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"Relax Prediction\",\"description\":\"Run structural relaxation after prediction\",\"default\":false,\"tooltip\":\"Relax the predicted structure using AMBER.\"},\"databases\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"maxItems\":3,\"minItems\":1},{\"type\":\"null\"}],\"title\":\"MSA Databases\",\"description\":\"Databases used for Multiple Sequence Alignment. By default, uniref90, mgnify, and small_bfd are used.Choice of databases(s) can significantly impact downstream structure prediction, so we recommend modifying carefully.\",\"default\":[\"uniref90\",\"mgnify\",\"small_bfd\"],\"tooltip\":\"Sequence databases used for Multiple Sequence Alignment.\"},\"contigs\":{\"type\":\"string\",\"maxLength\":100000,\"minLength\":0,\"title\":\"Contiguous Regions\",\"description\":\"Historically, contigs stands for 'contiguous [protein regions]'. This string defines a protein that is being generated. It is a specification written in a domain-specific language that tells RFdiffusion which part of the input protein are to be kept and what kind of a binder (or a scaffold) needs to be constructed. As an example, a string 'A10-100/0 50-150' instructs RFdiffusion to keep amino acids 10-100 in Chain A [from the input PDB file], then break the chain (special '/0' notation, which signifies the end of the chain and thus effectively makes 'A10-100' a new target protein), and construct a new chain (effectively a binder protein) of length 50 to 150 amino acids.\"},\"hotspot_res\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"maxItems\":5000,\"minItems\":0},{\"type\":\"null\"}],\"title\":\"Hotspot Residues\",\"description\":\"The hotspot residues string provides a way to specify which region the new protein (binder) must contact with the original input protein (a target), therefore we can guide a binder to a specific region. In UI, the format of the string is a comma-separated list of amino acids present in the input PDB file, e.g. 'A50,A51,A52' specifies three hotspot residues in the Chain A, at positions 50, 51 and 52. Note that in API, however, the amino acids are specified as a list of strings, e.g. ['A50', 'A51', ...].\"},\"diffusion_steps\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Number of diffusion steps\",\"description\":\"RFdiffusion is a diffusion generative model, it was trained by diffusing (adding noise) to a training data set. The generative process works by reversing the time steps (i.e. denoising): starting from randomly placed atoms, and reverse-diffusing the positions to arrive at a probable atom positions. The diffusion steps parameter tells RFdiffusion how many steps it will run the denoising process for. 1 is the minimum, 50 is the default.\",\"default\":50,\"max\":50,\"min\":1},\"random_seed\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Random Seed\",\"description\":\"The model allows users to set or not set the random seed based on the specific needs. For example, if reproducibility is crucial, it is recommended to set a fixed seed. However, for tasks requiring exploration and diversity, users might choose not to set the seed, allowing the model to leverage the benefits of randomness.\"},\"ca_only\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"CA-Only (alpha carbons) model enable flag\",\"description\":\"CA-only model helps to address specific needs in protein design where focusing on the alpha carbon (CA) atoms can be advantageous\",\"default\":false},\"use_soluble_model\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"Soluble model enable flag\",\"description\":\"ProteinMPNN offers both soluble and non-soluble models to cater to the specific needs of different protein design tasks. Soluble models are better suited for applications requiring high solubility, such as biotechnological processes, pharmaceutical development, and biochemical assays. Non-soluble models are advantageous for membrane protein studies, structural biology, and certain industrial applications where solubility is less critical or where proteins need to function in hydrophobic environments. This flexibility allows researchers to choose the appropriate model based on the specific requirements of their project.\",\"default\":false},\"num_seq_per_target\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Number of amino acid sequences to generate [per target protein]\",\"description\":\"This parameter specifies the number of sequences to generate per target protein structure. By setting num_seq_per_target, users can determine how many different sequences the model should predict that will fold into the given protein backbone structure.\",\"default\":1,\"max\":100,\"min\":1},\"sampling_temp\":{\"anyOf\":[{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"maxItems\":100,\"minItems\":1},{\"type\":\"null\"}],\"title\":\"Sampling temperatures\",\"description\":\"The units for sampling temperatures in ProteinMPNN are dimensionless and range from 0 to 1. This parameter is used to adjust the probability values for the 20 amino acids at each position in the sequence, thereby controlling the diversity of the design outcomes. Higher values lead to increased diversity in the designed results, while lower values result in less diversity and more conservative designs. Recommended range is from 0.1 to 0.3.\",\"default\":[0.1]}},\"type\":\"object\",\"required\":[\"sequence\",\"contigs\"],\"title\":\"ProteinDesignGenerationInputs\"},\"ProteinDesignGenerationOutputs\":{\"properties\":{\"result\":{\"type\":\"object\",\"title\":\"Results\",\"description\":\"A result dictionary containing results for each tool.\"},\"metrics\":{\"type\":\"object\",\"maxProperties\":200,\"minProperties\":0,\"title\":\"Metrics\",\"description\":\"Runtime metrics for the given tools.\"}},\"type\":\"object\",\"required\":[\"result\",\"metrics\"],\"title\":\"ProteinDesignGenerationOutputs\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T17:40:06.016Z\",\"nvcfFunctionId\":\"2e0014e9-0ceb-4401-b3e9-52536a5ffe71\",\"createdDate\":\"2025-01-13T17:26:05.637Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"blueprintUrl\":\"https://github.com/NVIDIA-BioNeMo-blueprints/generative-protein-binder-design\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: Your use of this API is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this blueprint is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false},\"artifactName\":\"protein-binder-design-for-drug-discovery\"},\"config\":{\"name\":\"protein-binder-design-for-drug-discovery\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"generative-virtual-screening-for-drug-discovery\",\"displayName\":\"Build A Generative Virtual Screening Pipeline\",\"publisher\":\"nvidia\",\"shortDescription\":\"This blueprint shows how generative AI and accelerated NIM microservices can design optimized small molecules smarter and faster.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/generative-virtual-screening-for-drug-discovery.jpg\",\"labels\":[\"BioNemo\",\"Blueprint\",\"Chemistry\",\"Docking\",\"NIM\",\"NVIDIA BioNemo\",\"Drug Discovery\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"NIM\",\"value\":\"openfold2,msa-search,genmol-generate,diffdock\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-08-22T20:54:17.545Z\",\"description\":\"$64\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-26T14:22:02.580Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"939259ed-f12a-47c5-89bd-51c3a3a5806b\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"Generative Virtual Screening\",\"version\":\"0.1.0\"},\"paths\":{\"/invoke\":{\"post\":{\"summary\":\"Design Binders\",\"operationId\":\"design_binders_invoke_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/GenerativeVirtualScreeningRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Responses\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}}}},\"/v2/health/ready\":{\"get\":{\"summary\":\"Health Check\",\"operationId\":\"health_check_v2_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}},\"/v1/health/ready\":{\"get\":{\"summary\":\"Health Check\",\"operationId\":\"health_check_v1_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{}}}}}}}},\"components\":{\"schemas\":{\"CADDSample\":{\"properties\":{\"ligand\":{\"type\":\"string\",\"title\":\"Ligand\"},\"ligand_score\":{\"type\":\"number\",\"title\":\"Ligand Score\"},\"ligand_score_type\":{\"type\":\"string\",\"title\":\"Ligand Score Type\"},\"best_position_confidence\":{\"type\":\"number\",\"title\":\"Best Position Confidence\"},\"docking_result\":{\"$ref\":\"#/components/schemas/DockingResult\"}},\"type\":\"object\",\"required\":[\"ligand\",\"ligand_score\",\"ligand_score_type\",\"best_position_confidence\",\"docking_result\"],\"title\":\"CADDSample\"},\"DockingResult\":{\"properties\":{\"ligand_positions\":{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"title\":\"Ligand Positions\"},\"position_confidence\":{\"items\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}]},\"type\":\"array\",\"title\":\"Position Confidence\"},\"ligand\":{\"type\":\"string\",\"title\":\"Ligand\"},\"trajectory\":{\"title\":\"Trajectory\"}},\"type\":\"object\",\"required\":[\"ligand_positions\",\"position_confidence\",\"ligand\"],\"title\":\"DockingResult\"},\"GenerativeVirtualScreeningRequest\":{\"properties\":{\"sequence\":{\"type\":\"string\",\"minLength\":1,\"pattern\":\"^[ARNDCQEGHILKMFPSTWYV]+\",\"title\":\"Input Polypeptide Sequence\",\"description\":\"An input polypeptide (i.e., amino acid) sequence that must be composed of valid Amino Acid IUPAC symbols.\",\"tooltip\":\"The input sequence for structural prediction.\"},\"e_value\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}],\"title\":\"MSA e-value.\",\"description\":\"The e-value used for filtering hits when building the Multiple Sequence Alignment.\",\"default\":0.0001,\"tooltip\":\"Sequence e-value for filtering sequences seeding the MSA.\"},\"iterations\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"MSA Iterations\",\"description\":\"The number of MSA iterations to perform.\",\"default\":1,\"tooltip\":\"The number of iterations performed by the MSA algorithm.\"},\"databases\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"maxItems\":3,\"minItems\":1},{\"type\":\"null\"}],\"title\":\"MSA Databases\",\"description\":\"Databases used for Multiple Sequence Alignment. By default, uniref90, mgnify, and small_bfd are used.Choice of databases(s) can significantly impact downstream structure prediction, so we recommend modifying carefully.\",\"default\":[\"uniref90\",\"mgnify\",\"small_bfd\"],\"tooltip\":\"Sequence databases used for Multiple Sequence Alignment.\"},\"run_structural_template_search\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"Run structural templating\",\"description\":\"Run structural templating\",\"default\":false,\"tooltip\":\"If true, templating is performed.\"},\"smiles\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Molecule SMILES\"},\"num_molecules\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":1000,\"minimum\":1},{\"type\":\"null\"}],\"title\":\"Number of Molecules to Generate\",\"default\":30},\"temperature\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":10,\"minimum\":0.01},{\"type\":\"null\"}],\"title\":\"Temperature Scaling Factor for Softmax Sampling\",\"default\":1},\"noise\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":2,\"minimum\":0},{\"type\":\"null\"}],\"title\":\"Noise Factor for Top-K Sampling\",\"default\":1},\"step_size\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":10,\"minimum\":1},{\"type\":\"null\"}],\"title\":\"Diffusion Step Size\",\"default\":1},\"scoring\":{\"anyOf\":[{\"$ref\":\"#/components/schemas/ScoringMethod\"},{\"type\":\"null\"}],\"title\":\"Scoring Method\",\"default\":\"QED\"},\"num_poses\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":100,\"minimum\":1},{\"type\":\"null\"}],\"title\":\"Molecule SMILES\",\"default\":20}},\"type\":\"object\",\"required\":[\"sequence\"],\"title\":\"GenerativeVirtualScreeningRequest\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Responses\":{\"properties\":{\"protein\":{\"type\":\"string\",\"title\":\"Protein\"},\"ligands\":{\"items\":{\"$ref\":\"#/components/schemas/CADDSample\"},\"type\":\"array\",\"title\":\"Ligands\"}},\"type\":\"object\",\"required\":[\"protein\",\"ligands\"],\"title\":\"Responses\"},\"ScoringMethod\":{\"type\":\"string\",\"enum\":[\"QED\",\"LogP\"],\"title\":\"ScoringMethod\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-26T14:22:03.222Z\",\"nvcfFunctionId\":\"eac48986-5e3c-41e0-b79e-98b3c523c466\",\"createdDate\":\"2024-08-22T20:54:17.954Z\",\"attributes\":{\"apiDocsUrl\":\"FILL_THIS IN\",\"blueprintUrl\":\"https://github.com/NVIDIA-NIM-Agent-Blueprints/generative-virtual-screening\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: Your use of this API is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this blueprint is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false},\"artifactName\":\"generative-virtual-screening-for-drug-discovery\"},\"config\":{\"name\":\"generative-virtual-screening-for-drug-discovery\",\"type\":\"blueprint\"}}],\"items\":[\"$65\",\"$66\",\"$67\",\"$68\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"Accelerated Computing for Digital Biology\"}],[\"$\",\"$L26\",null,{\"href\":\"/blueprints?filters=blueprintType%3Ablueprinttype_nvidia_bionemo\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"AI-driven drug discovery and accelerated genomics workflows.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"69:Te52,"])</script><script>self.__next_f.push([1,"## Experience Your Personal Research Assistant\n\nThis experience shows how an AI research agent, informed by many data sources, can synthesize hours of research in minutes. The AI-Q NVIDIA Blueprint enables developers to build AI agents that use reasoning and connect to many data sources and tools to distill in-depth source materials with efficiency and precision. Using AI-Q, agents summarize large data sets, generating tokens 5x faster and ingesting petabyte scale data 15x faster with better semantic accuracy. \n\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/aiq/diagram.jpg)\n\n## Key Features\n- Advanced semantic query \n - Multimodal PDF data extraction and retrieval with NVIDIA NeMo Retriever\n - 15x faster ingestion of enterprise data\n - 3x lower retrieval latency \n - Multilingual and cross-lingual \n - Reranking to further improve accuracy\n - GPU-accelerated index creation and search\n- Fast reasoning\n - Llama Nemotron reasoning capabilities delivering the highest accuracy and lowest latency for analyzing datasets, identifying patterns, and proposing solutions\n\n## Software used in this blueprint\n\n**NVIDIA Technology**\n- [llama-3.3-nemotron-49b-instruct](https://build.nvidia.com/nvidia/llama-3_3-nemotron-49b-instruct)\n- [llama-3.2-nv-embedqa-1b-v2](https://build.nvidia.com/nvidia/llama-3_2-nv-embedqa-1b-v2)\n- [llama-3.2-nv-rerankqa-1b-v2](https://build.nvidia.com/nvidia/llama-3_2-nv-rerankqa-1b-v2)\n- [nemoretriever-graphic-elements-v1](https://build.nvidia.com/nvidia/nemoretriever-graphic-elements-v1)\n- [nemoretriever-table-structure-v1](https://build.nvidia.com/nvidia/nemoretriever-table-structure-v1)\n- [nemoretriever-page-elements-v2](https://build.nvidia.com/nvidia/nemoretriever-page-elements-v2)\n- [nemoretriever-parse](https://build.nvidia.com/nvidia/nemoretriever-parse)\n- [paddleocr](https://build.nvidia.com/baidu/paddleocr)\n- [llama-3.1-nemotron-70b-instruct](https://build.nvidia.com/nvidia/llama-3_1-nemotron-70b-instruct)\n\n**3rd Party Software**\n- [Tavily](https://tavily.com/)\n- [LangChain](https://www.langchain.com/)\n- Milvus database (accelerated with NVIDIA [cuVS](https://github.com/rapidsai/cuvs))\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: Governing Terms: Your use of this trial experience is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Additional Information: [LLAMA 3.3 COMMUNITY LICENSE AGREEMENT](https://www.llama.com/llama3_3/license/); [LLAMA 3.2 COMMUNITY LICENSE AGREEMENT](https://www.llama.com/llama3_2/license/); [LLAMA 3.1 COMMUNITY LICENSE AGREEMENT](https://www.llama.com/llama3_1/license/). Built with Llama."])</script><script>self.__next_f.push([1,"6a:T1c6c,"])</script><script>self.__next_f.push([1,"Today's large language models (LLMs) are subject to a trade-off between reasoning capabilities and computational efficiency. While powerful models excel at complex reasoning tasks, sophisticated test-time compute, and level 2 thinking (reasoning about their own reasoning), they’re computationally expensive and slower, making them impractical for simpler tasks. The NVIDIA AI Blueprint for an LLM router is designed to mitigate this trade-off by intelligently directing prompts to the most appropriate model, ensuring optimal balance between reasoning depth and computational efficiency. Through its lightweight classification models that run in milliseconds, it routes simple queries to fast, efficient models and directs prompts that demand careful analysis and self-reflective reasoning to more powerful models that can apply extensive test-time computation.\n\nThe blueprint achieves this through a flexible architecture that supports multiple routing strategies, from task-based classification to user-intent analysis to reasoning-based routing. Using specialized classification models, it analyzes each prompt for complexity, required domain knowledge, and need for iterative thinking, enabling organizations to maintain high-quality responses for complex reasoning tasks while optimizing computational resources. This strategic routing lets organizations scale their AI systems efficiently and ensure deep reasoning capabilities are available when needed, fundamentally transforming how we deploy and utilize language models in production environments.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/llm-router/diagram.jpg)\n\n## What’s Included in the Blueprint\n\n### Key Features\nThis reference architecture includes an architectural diagram, an NVIDIA Brev launchable with a Jupyter notebook for rapid exploration and experimentation, and source code for local deployment and customization. The LLM Router supports the following key features and components:\n\n- **Low-Latency Router Controller:** To manage routing logic and decision-making for optimal query distribution. The router models are very small, so they don’t add extra latency and are easily fine-tunable.\n- **Response Evaluation Strategies:** For assessing LLM outputs to improve routing accuracy and decisions.\n- **Multi-LLM Routing:** For routing among multiple models, going beyond binary selection.\n- **Customization Workflows:** For fine-tuning router models for specific use cases.\n- **Flexible Routing Methodologies:** For cost-based, response quality-based, task-based, and intent-based routing.\n- **Modular Design:** Can deploy either the router controller with the router server or only the router server alone and use a different proxy. The router controller and router server can be deployed in separate systems. The router server needs a GPU and the router controller doesn't need a GPU.\n- **Powerful Abstraction Layer:** Streamlines deployment by handling routing pipelines and model orchestration behind the scenes.\n\n## Software Used in This Blueprint\n\n**NVIDIA NIM™ microservices**\n\n- [Llama 3.1 8B Instruct](https://build.nvidia.com/meta/llama-3_1-8b-instruct)\n- [Llama 3.1 70B Instruct](https://build.nvidia.com/meta/llama-3_1-70b-instruct)\n- [Mixtral 8x22B Instruct](https://build.nvidia.com/mistralai/mixtral-8x22b-instruct)\n- [DeepSeek R1](https://build.nvidia.com/deepseek-ai/deepseek-r1)\n\n**Other**\n\n- [NVIDIA Triton Inference Server](https://developer.nvidia.com/triton-inference-server)\n\n## Minimum System Requirements\n\n### Hardware Requirements\n- Any NVIDIA GPU with an NVIDIA architecture newer than Volta™ (V100), such as Turing™ (T4), Ampere™ (A100, RTX 30 series), Hopper™ (H100), or later.\n\n### Software Requirements\n- [Git LFS](https://git-lfs.com/)\n- [Docker](https://www.docker.com/)\n- Docker Compose\n- NVIDIA API key from [build.nvidia.com](http://build.nvidia.com) (see [instructions](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html#option-1-from-api-catalog))\n\n## Ethical Considerations\nNVIDIA believes trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety and Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n\n\u003e **Warning:** The Terms of Use section below is a work in progress and will be updated with the final terms.\n\n## Terms of Use\nGOVERNING TERMS: The software is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [Product-Specific Terms for NVIDIA AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/). Use of the Complexity and Task Qualifier model is governed by the [NVIDIA Open Model License Agreement](https://developer.download.nvidia.com/licenses/nvidia-open-model-license-agreement-june-2024.pdf). Additional Information: [MIT License](https://github.com/microsoft/DeBERTa/blob/master/LICENSE).\n\n#### Meta Llama 3.1 8B, Llama 3.1 70B Instruct\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/);\n\n#### Mixtral 8x22B Instruct\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/);\n\n#### DeepSeek R1\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\n\nUse of these model is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.1 Community License Agreement, Built with Llama;"])</script><script>self.__next_f.push([1,"6b:T11f1,"])</script><script>self.__next_f.push([1,"Unlock the power of on-the-go learning and tackle the challenge of information overload with generative AI-powered audio read-outs. Use this blueprint to build a generative AI application that transforms PDF data—such as training documents, technical research, or documentation—into personalized audio content. \n\nLeverage large language models (LLMs), text-to-speech, and NVIDIA NIM microservices to deploy a customized solution tailored to your organization’s proprietary data. This approach remains compliant with privacy requirements throughout the process. \n\nThis blueprint is flexible and customizable, so you can add additional functionality that suits your users’ needs, whether that is specific branding, analytics, real-time translation, or a digital human interface to deepen engagement. \n\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/pdf-to-podcast/diagram.jpg)\n\n## Key Features\n**PDF to Markdown Service**\n- Extracts content from PDFs and converts it into markdown format for further processing.\n\n**Monologue or Dialogue Creation Service**\n- AI processes markdown content, enriching or structuring it to create natural and engaging audio content.\n\n**Text-to-Speech (TTS) Service**\n- Converts the processed content into high-quality speech.\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [Llama 3.1 8B Instruct](https://build.nvidia.com/meta/llama-3_1-8b-instruct)\n- [Llama 3.1 70B Instruct](https://build.nvidia.com/meta/llama-3_1-70b-instruct)\n- [Llama 3.1 405B Instruct](https://build.nvidia.com/meta/llama-3_1-405b-instruct)\n\n**3rd Party Software**\n\n- [Langchain](https://www.langchain.com/)\n- [Docling](https://github.com/DS4SD/docling) Document Parser for PDF to Markdown Service\n- [ElevenLabs](https://elevenlabs.io/) for Text-to-Speech Service\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.3-70B\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\n#### Meta Llama 3.1 8B, Llama 3.1 70B Instruct, Llama 3.1 405B Instruct\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\nUse of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.1 Community License Agreement, Built with Llama."])</script><script>self.__next_f.push([1,"6c:T15b2,"])</script><script>self.__next_f.push([1,"The NVIDIA AI Blueprint for RAG gives developers a foundational starting point for building scalable, customizable retrieval pipelines that deliver both high accuracy and throughput. Use this blueprint to create RAG applications that provide context-aware responses by connecting LLMs to extensive multimodal enterprise data—an essential capability for most generative AI use cases. \n\nThis blueprint can be utilized as-is, combined with other NVIDIA Blueprints, such as the Digital Human Blueprint or the AI Virtual Assistant for customer service, or integrated with an agent to support more advanced use cases. Get started with this reference architecture to unlock actionable insights, ground your decisions in relevant data, and boost overall productivity.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/build-an-enterprise-rag-pipeline/diagram.jpg)\n\n## Key Features\n* Multimodal data extraction support with text, tables, charts, and infographics \n* Hybrid search with dense and sparse search \n* Opt-in image captioning with vision language models (VLMs) \n* Reranking to further improve accuracy \n* GPU-accelerated Index creation and search \n* Multi-turn conversations \n* Multi-session support \n* Telemetry and observability \n* Opt-in for reflection to improve accuracy \n* Opt-in for guardrailing conversations \n* Sample user interface \n* OpenAI-compatible APIs \n* Decomposable and customizable\n\n## Minimum System Requirements\n\n**Hardware Requirements**\nThe blueprint offers two primary modes of deployment. By default, it deploys the referenced NIM microservices locally. Each method lists its minimum required hardware. This will change if the deployment turns on optional configuration settings.\n\n* Docker \n * 4xH100 or 6xA100 \n* Kubernetes \n * 9xH100 or 11XA100 \n* The blueprint provides the alternative to use NGC-hosted models, in which case one GPU will be required to host the [NVIDIA cuVS](https://developer.nvidia.com/cuvs)\\-accelerated vector database. \n* The blueprint can be modified to use additional NIM microservices hosted by NVIDIA. \n\n**OS Requirements**\n- Ubuntu 22.04 OS\n\n**Deployment Options**\n- Docker\n- Kubernetes\n\n## Software used in this blueprint\n\n**NVIDIA Technology**\n* [NeMo Retriever Llama 3.2 Embedding NIM](https://build.nvidia.com/nvidia/llama-3_2-nv-embedqa-1b-v2) \n* [NeMo Retriever Llama 3.2 Reranking NIM](https://build.nvidia.com/nvidia/llama-3_2-nv-rerankqa-1b-v2) \n* [Llama 3.1 70B Instruct NIM](https://build.nvidia.com/meta/llama-3_1-70b-instruct) \n* [NeMo Retriever Page Elements NIM](https://build.nvidia.com/nvidia/nemoretriever-page-elements-v2) \n* [NeMo Retriever Table Structure NIM](https://build.nvidia.com/nvidia/nemoretriever-table-structure-v1) \n* [NeMo Retriever Graphic Elements NIM](https://build.nvidia.com/nvidia/nemoretriever-graphic-elements-v1) \n* [PaddleOCR NIM](https://build.nvidia.com/baidu/paddleocr) \n* [NeMo Retriever Parse NIM](https://build.nvidia.com/nvidia/nemoretriever-parse) *(optional)* \n* [Llama 3.1 NemoGuard 8B Content Safety NIM](https://build.nvidia.com/nvidia/llama-3_1-nemoguard-8b-content-safety) *(optional)* \n* [Llama 3.1 NemoGuard 8B Topic Control NIM](https://build.nvidia.com/nvidia/llama-3_1-nemoguard-8b-topic-control) *(optional)* \n* [Llama 3.2 11B Vision Instruct NIM](https://build.nvidia.com/meta/llama-3.2-11b-vision-instruct) *(optional)* \n* [Mixtral 8x22B Instruct 0.1](https://build.stg.ngc.nvidia.com/mistralai/mixtral-8x22b-instruct) *(optional)* \n\n**3rd Party Software**\n* [LangChain](https://www.langchain.com/) \n* Milvus database (accelerated with [NVIDIA **cuVS**)](https://github.com/rapidsai/cuvs)\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nThis blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/). The models are governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/) and the [NVIDIA RAG dataset](https://github.com/NVIDIA-AI-Blueprints/rag/tree/v2.0.0/data/multimodal) which is governed by the [NVIDIA Asset License Agreement](https://github.com/NVIDIA-AI-Blueprints/rag/blob/main/data/LICENSE.DATA).\n\nThe following models that are built with Llama are governed by the [Llama 3.2 Community License Agreement](https://www.llama.com/llama3_2/license/): llama-3.1-70b-instruct, nvidia/llama-3.2-nv-embedqa-1b-v2, and nvidia/llama-3.2-nv-rerankqa-1b-v2."])</script><script>self.__next_f.push([1,"6d:T1302,"])</script><script>self.__next_f.push([1,"The integration of CrewAI with NVIDIA NIM is demonstrated with a blueprint for building AI agents specialized in code documentation.\n\nThis blueprint showcases the flexibility of CrewAI when combined with NVIDIA to solve real world challenges, such as improving and automating software documentation processes.\n\nThis use case addresses critical issues such as inconsistent documentation processes, and maintenance challenges. By leveraging the power of CrewAI and NVIDIA NIM, teams can enhance productivity, minimize confusion and streamline the creation and maintenance of high-quality software documentation. Developers can use this flexible reference blueprint to update an existing CrewAI solution with NVIDIA AI, create new software documentation agent, or apply it to a different use case that includes CrewAI and NVIDIA.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/code-documentation-for-software-development/diagram.jpg)\n\n## Key Features\nThis reference architecture leverages CrewAI and **Llama 3.3-70B LLM NIM** and **NeMo Retriever E5 embedding NIM** as its underlying LLM and embedding model respectively, to generate comprehensive, high-quality documentation for GitHub repositories. The system employs a multi-agent workflow divided into two key stages:\n\n### Ingestion Phase\n- **WebsiteSearchTool:** This tool is used to embed and index mermaid examples from mermaid.js.org website using NVIDIA NeMo Retriever E5 embedding NIM.\n\n### Agent Flow\n1. **Codebase Analysis and Strategy Planning:**\n - **Analyze Codebase:** Planner agents inspect the repository to map its structure, identify key components, and understand interdependencies.\n - **Develop Strategy:** They create a tailored documentation plan based on the analysis.\n\n2. **Documentation Creation and Review:**\n - **High-Level Documentation:** One agent generates clear, comprehensive documentation introducing the project and its architecture.\n - **Quality Assurance:** Another agent ensures accuracy, consistency, and completeness across all documentation.\n\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [Llama 3.3 70B NVIDIA NIM](https://build.nvidia.com/meta/llama-3_3-70b-instruct)\n- [NeMo Retriever E5 embedding NIM](https://build.nvidia.com/nvidia/nv-embedqa-e5-v5)\n\n**3rd-Party Technologies**\n\n- [CrewAI](https://www.crewai.com/)\n\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.3 70B\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\n#### NVIDIA Retrieval QA E5 Embedding Model\nUse of this model is governed by [MIT license](https://opensource.org/license/MIT).\n\nUse of these models is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.3 Community License Agreement, Built with Llama."])</script><script>self.__next_f.push([1,"6e:T1159,"])</script><script>self.__next_f.push([1,"Traditional RAG systems are effective at answering individual questions but have limited utility for organizations requiring more comprehensive insights and decision-making content. This Structured Report Generation blueprint with LangGraph addresses this need by using NVIDIA NIM microservices to create structured, decision-oriented reports based on user-supplied topics and report structure.\n\nUnlike other Structured Report Generation agents, this blueprint shows how NVIDIA NIM microservices with LangChain improves scalability of data sources, accuracy and throughput. Developers can use this flexible reference blueprint to update an existing LangChain solution with NVIDIA AI, create a new structured report generation agent or apply it to a different use case that includes LangChain and NVIDIA NIM.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/structured-report-generation/diagram.jpg)\n\n## Key Features\nWithin this blueprint, LangChain's LangGraph is used to build a Report Planning Agent that takes in user defined topics and structure, then plans the topics of the section indicated in the structure. The Research Agent uses Tavily to do web search on the given topics and the Report Writing Agent uses this information to write the sections and synthesize the final report.\n\nA two-phase approach is used for **planning** and **research**:\n\nPhase 1 - Planning\n- Analyzes user inputs\n- Maps out report sections\n\nPhase 2 - Research\n- Conducts parallel web research via Tavily API\n- Processes relevant data for each section\n\nThe report is then **written** in strategic sequence:\n1. Write research-based sections in parallel\n2. Write introductions, conclusions, and connect each of the sections\n\nAll sections maintain awareness of each other's content for consistency.\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [Llama 3.3 70B NVIDIA NIM](https://build.nvidia.com/meta/llama-3_3-70b-instruct)\n\n**3rd-Party Technologies**\n\n- [LangChain LangGraph](https://www.langchain.com/langgraph)\n- [Tavily](https://www.tavily.com/)\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.3 70B\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\nUse of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.3 Community License Agreement, Built with Llama."])</script><script>self.__next_f.push([1,"6f:T154e,"])</script><script>self.__next_f.push([1,"This blueprint demonstrates how to build an AI agent that automates blog post creation by leveraging LlamaIndex's document processing and agentic framework capabilities alongside NVIDIA Llama 3.3 70B LLM NIM and NeMo Retriever Llama3.2 embedding model. The agent handles the entire writing process - from research and outlining to drafting and editing - to produce high-quality, well-researched blog content.\n\nThe system provides an automated workflow that orchestrates the entire content creation process - from initial research and outlining to drafting and quality assurance - while maintaining high standards of accuracy and comprehensiveness through the integration of LlamaIndex's document processing with NVIDIA's powerful language and retrieval models. Developers can use this flexible reference blueprint to update an existing LlamaIndex solution with NVIDIA AI, create a new blog creation agent, or apply it to a different use case that includes LlamaIndex and NVIDIA.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/document-research-assistant-for-blog-creation/diagram.jpg)\n\n## Key Features\n### Ingestion Phase\n1. Document parsing: Use LlamaParse to convert PDFs into Markdown.\n2. Document ingestion: Employ NeMo Retriever Llama3.2 embedding model to ingest the documents into a vector database.\n3. Query engine instantiation: Leverage LlamaIndex to create a query engine that can answer questions about the documents.\n\n### Query Phase\n1. Research Request: User provides a set of tools (e.g., a query engine with data about San Francisco's budget) and a content request (e.g., a question for a blog post).\n2. Outline Generation Agent: Deploys an agent to structure the blog post into an actionable outline.\n3. Question Generation Agent: Generates a list of questions necessary to address the outline effectively, and breaks the questions into discrete units that can be answered concurrently, using available tools for data collection.\n4. Content Generation Agent: A writer agent synthesizes the gathered answers into a cohesive blog post.\n5. Critic Agent: reviews the content for accuracy, coherence, and completeness, determining if revisions are necessary.\n6. Iterative Refinement: If improvements are needed, the workflow repeats by generating additional questions and gathering more information until the desired quality is reached.\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [Llama 3.3 70B NVIDIA NIM](https://build.nvidia.com/meta/llama-3_3-70b-instruct)\n- [NeMo Retriever Llama3.2 embedding model](https://build.nvidia.com/nvidia/llama-3_2-nv-embedqa-1b-v2)\n\n**3rd-Party Technologies**\n\n- [LlamaIndex](https://www.llamaindex.ai/)\n\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.3 70B\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\nUse of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.3 Community License Agreement, Built with Llama.\n\n#### llama-3.2-nv-embedqa-1b-v2\nThe use of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/) and Llama 3.2 is licensed under the [Llama 3.2 Community License](https://www.llama.com/llama3_2/license/), Copyright © Meta Platforms, Inc. \nAll Rights Reserved."])</script><script>self.__next_f.push([1,"70:T12ea,"])</script><script>self.__next_f.push([1,"This blueprint utilizes Pipecat to create a deployable voice agent, built on NVIDIA NIM microservices, for seamless integration into production environments.\nA production-ready conversational voice agent requires the integration of several complex components, including multiple AI models (such as STT, LLM, TTS, and guardrails), conversation context management, and frameworks for state management and legacy system integration. Additionally, it involves handling hooks for RAG, phrase endpointing, interruption management, ultra-low latency network transport, echo cancellation, and background noise reduction. The solution also requires integration with telephony systems, client-side SDKs for connection management and multimedia exchange, and integration with evaluation and observability tools. All these elements must be managed to ensure conversational latency (500-1500ms for voice-to-voice responses). \n\nPipecat, created by Daily.co, is an open-source framework that addresses these challenges, supporting 40+ AI models and services as plugins and offering SDKs for various platforms including Python, JavaScript, React, iOS, Android, and C++.\n\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/voice-agent-framework-for-conversational-ai/diagram.jpg)\n\n## Key Features\nThis blueprint gives developers a one-click deployable conversational voice agent. Enterprise easily can build and deploy voice agents across use cases, including customer service, virtual assistants, productivity, gaming, and IoT.\n\nThe blueprint is:\n- A starter kit for an enterprise-ready voice AI agent infrastructure solution\n- An accessible entry point for developers who want to learn about voice AI\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [Llama 3.3 70B NVIDIA NIM](https://build.nvidia.com/meta/llama-3_3-70b-instruct)\n- [NVIDIA Riva Automatic Speech Recognition (ASR) NIM](https://build.nvidia.com/nvidia/parakeet-ctc-1_1b-asr)\n- [NVIDIA Riva Text to Speech (TTS) NIM](https://build.nvidia.com/nvidia/fastpitch-hifigan-tts)\n\n\n**3rd-Party Technologies**\n\n- [Pipecat](https://github.com/daily-co/nimble-pipecat)\n\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint are governed by the NVIDIA AI Foundation Models [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.3 70B\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and the [Product Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); \n\nUse of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](\u003chttps://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C\u003e). ADDITIONAL INFORMATION: Llama 3.3 Community License Agreement, Built with Llama.\n\nNVIDIA Riva Models\nPlease refer to the Governing terms for NVIDIA parakeet-ctc-1_1b-asr [here](https://build.nvidia.com/nvidia/parakeet-ctc-1_1b-asr/modelcard)\nPlease refer to the Governing terms for NVIDIA FastPitch-HifiGAN [here](https://build.nvidia.com/nvidia/fastpitch-hifigan-tts/modelcard)"])</script><script>self.__next_f.push([1,"71:T153c,"])</script><script>self.__next_f.push([1,"This blueprint extends the existing NVIDIA AI Blueprint - [AI Virtual Assistant for Customer Service](https://build.nvidia.com/nvidia/ai-virtual-assistant-for-customer-service) - to demonstrate how traceability can be added to an AI Agent. \n\nWeights \u0026 Biases Weave framework helps customers evaluate, monitor, and iterate on their AI applications to accelerate the development and deployment process. Weave enables continuous improvement in quality, latency, cost, and safety by running comprehensive evaluations, keeping pace with new models, debugging, and monitoring production performance—all while ensuring secure collaboration. Any enterprise wanting to take a generative AI application from pilot to production needs to have a way to monitor, evaluate and iterate for gaining insights on how their application is performing and to further power the data flywheel.\n\nDevelopers can use this reference blueprint to extend W\u0026B Weave capabilities to the AI Virtual Assistant for Customer Service blueprint or apply it to another NVIDIA AI Blueprint.\n\n## Architecture Diagram\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/traceability-for-agentic-ai/diagram.jpg)\n\n## Key Features\nThis blueprint extension achieves the following:\n- Showcases how Weights \u0026 Biases integrates into the workflow to provide seamless tracing, evaluations, and iteration tooling, ensuring a more efficient and accelerated iteration and promotion process\n- Demonstrates how to bring an AI application closer to production readiness by adding Weave from Weights and Biases\n\n\n## Minimum System Requirements\nThe solution leverages NVIDIA's cloud-based API Catalog endpoints, eliminating the need for local GPU hardware. All model inference is performed on NVIDIA's cloud infrastructure.\n\n## Software used in this blueprint\n**NIM microservices**\n\n- [NVIDIA NeMo Retriever embedding NIM](https://build.nvidia.com/nvidia/nv-embedqa-e5-v5)\n- [NVIDIA NeMo Retriever Mistral 4B reranking NIM](https://build.nvidia.com/nvidia/nv-rerankqa-mistral-4b-v3)\n- [Llama 3.1 70B instruct NIM](https://build.nvidia.com/meta/llama-3_1-70b-instruct)\n- [Nemotron-4 340B NIM](https://build.nvidia.com/nvidia/nemotron-4-340b-instruct)\n\nPlease refer to [AI Virtual Assistant for Customer Service](https://build.nvidia.com/nvidia/ai-virtual-assistant-for-customer-service) for the details on the foundational blueprint.\n\n**3rd-Party Technologies**\n\n- [W\u0026B Weave](https://wandb.ai/site/weave/)\n\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n\n## License\nUse of the models in this blueprint is governed by the [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf).\n\n## Terms of Use\nGOVERNING TERMS: The blueprint is governed by the [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/).\n \n \n#### Meta Llama 3.1 70B Instruct\nGOVERNING TERMS: The NIM container is governed by the NVIDIA Software License Agreement and the Product Specific Terms for AI Products;\n\nUse of this model is governed by the NVIDIA AI Foundation Models Community License Agreement. ADDITIONAL INFORMATION: Llama 3.1 Community License Agreement, Built with Llama.\n\n#### NVIDIA Retrieval QA E5 Embedding Model\nGOVERNING TERMS: The NIM container is governed by [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); and the use of this model is governed by the ai-foundation-models-community-license.pdf (nvidia.com). ADDITIONAL INFORMATION: MIT license.\n\n\n#### NeMo Retriever QA Mistral 4B Reranking v3\nGOVERNING TERMS: The NIM container is governed by [NVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [NVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); and the use of this model is governed by the [ai-foundation-models-community-license.pdf](https://docs.nvidia.com/ai-foundation-models-community-license.pdf) (nvidia.com). ADDITIONAL INFORMATION: Apache license."])</script><script>self.__next_f.push([1,"72:{\"name\":\"aiq\",\"type\":\"blueprint\"}\n73:{\"name\":\"llm-router\",\"type\":\"blueprint\"}\n74:{\"name\":\"pdf-to-podcast\",\"type\":\"blueprint\"}\n75:{\"name\":\"build-an-enterprise-rag-pipeline\",\"type\":\"blueprint\"}\n76:{\"name\":\"code-documentation-for-software-development\",\"type\":\"model\"}\n77:{\"name\":\"structured-report-generation\",\"type\":\"model\"}\n78:{\"name\":\"document-research-assistant-for-blog-creation\",\"type\":\"model\"}\n79:{\"name\":\"voice-agent-framework-for-conversational-ai\",\"type\":\"model\"}\n7a:{\"name\":\"traceability-for-agentic-ai\",\"type\":\"model\"}\n"])</script><script>self.__next_f.push([1,"38:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"aiq\",\"displayName\":\"Build an AI Agent for AI Research and Reporting\",\"publisher\":\"nvidia\",\"shortDescription\":\"Create AI agents that reason, plan, reflect and refine to produce high-quality reports based on source materials of your choice.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/aiq.jpg\",\"labels\":[\"Blueprint\",\"Llama Nemotron\",\"NIM\",\"NeMo Retriever\",\"Reasoning\",\"Retrieval-Augmented Generation\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:22:53.669Z\",\"description\":\"$69\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:24:26.940Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"f30743b4-a707-4da9-b832-dde3b86a9a38\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:24:27.350Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-18T19:22:53.966Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"blueprintUrl\":\"https://github.com/NVIDIA-AI-Blueprints/rag/tree/v1.0.0\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement\u003c/a\u003e and \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Notify Me When Available\",\"url\":\"https://www.nvidia.com/en-us/ai-data-science/generative-ai/news/\"}},\"artifactName\":\"aiq\"},\"config\":{\"name\":\"aiq\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"llm-router\",\"displayName\":\"LLM Router\",\"publisher\":\"nvidia\",\"shortDescription\":\"Route LLM requests to the best model for the task at hand.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llm-router.jpg\",\"labels\":[\"Blueprint\",\"LLM Router\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-04T01:51:54.970Z\",\"description\":\"$6a\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-17T21:58:09.756Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"8f2ea6ae-7335-429c-9c81-18f95386eacb\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-17T21:58:10.194Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-03-04T01:51:55.282Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2tp7rYvbzMjlRDOuPrVXZdpC2nI\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-AI-Blueprints/llm-router\"}},\"artifactName\":\"llm-router\"},\"config\":{\"name\":\"llm-router\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"pdf-to-podcast\",\"displayName\":\"PDF to Podcast\",\"publisher\":\"nvidia\",\"shortDescription\":\"Transform PDFs into AI podcasts for engaging on-the-go audio content.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/pdf-to-podcast.jpg\",\"labels\":[\"Conversational AI\",\"Multi-modal\",\"PDF-to-Podcast\",\"Text-to-speech\",\"blueprint\",\"NVIDIA AI\",\"AI Agent\",\"Text-to-Speech\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:40:07.235Z\",\"description\":\"$6b\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-16T13:52:26.704Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"876eaded-3748-4986-9545-2948c4091c0a\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-16T13:52:27.156Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-01-07T04:40:07.475Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2qxKCl3TbK61YrTIEKpWWmOnc8w\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-AI-Blueprints/pdf-to-podcast\"}},\"artifactName\":\"pdf-to-podcast\"},\"config\":{\"name\":\"pdf-to-podcast\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"build-an-enterprise-rag-pipeline\",\"displayName\":\"Build an Enterprise RAG pipeline\",\"publisher\":\"nvidia\",\"shortDescription\":\"Connect AI applications to multimodal enterprise data with a retrieval augmented generation (RAG) pipeline.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/build-an-enterprise-rag-pipeline.jpg\",\"labels\":[\"Blueprint\",\"NIM\",\"NeMo retriever\",\"Retrieval-Augmented Generation\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"NIM\",\"value\":\"llama-3_1-70b-instruct,llama-3_2-nv-embedqa-1b-v2,llama-3_2-nv-rerankqa-1b-v2,nemoretriever-page-elements-v2,nemoretriever-table-structure-v1,nemoretriever-graphic-elements-v1,paddleocr\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-24T04:02:23.879Z\",\"description\":\"$6c\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:35:58.984Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"194d36ae-6e73-4bda-b385-113303a9f06b\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:35:59.422Z\",\"nvcfFunctionId\":\"2aeb01be-7115-47e9-a676-8b04ca86e5e2\",\"createdDate\":\"2025-01-24T04:02:24.199Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"blueprintUrl\":\"https://github.com/NVIDIA-AI-Blueprints/rag/tree/v1.0.0\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Agreements | Enterprise Software | NVIDIA Software License Agreement\u003c/a\u003e and \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Agreements | Enterprise Software | Product Specific Terms for AI Product\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://brev.nvidia.com/launchable/deploy?launchableID=env-2uFJF22Hp8UdIVw0jRbHHYaixXx\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/NVIDIA-AI-Blueprints/rag/releases/tag/v2.0.0\"}},\"artifactName\":\"build-an-enterprise-rag-pipeline\"},\"config\":{\"name\":\"build-an-enterprise-rag-pipeline\",\"type\":\"blueprint\"}},{\"endpoint\":{\"artifact\":{\"name\":\"code-documentation-for-software-development\",\"displayName\":\"Code Documentation for Software Development\",\"publisher\":\"crewai\",\"shortDescription\":\"Document your github repositories with AI Agents using CrewAI and Llama3.3 70B NIM.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/code-documentation-for-software-development.jpg\",\"labels\":[\"AI Agents\",\"Blueprint\",\"Code Documentation\",\"CrewAI\",\"Partner\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"},{\"key\":\"PUBLISHERLOGO\",\"value\":\"https://assets.ngc.nvidia.com/products/api-catalog/code-documentation-for-software-development/publisher.png\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:26:00.974Z\",\"description\":\"$6d\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:27:29.920Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"6b1e77ec-2b84-4f0c-a277-18a63141411c\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:27:30.249Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-01-07T04:26:01.235Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"publisherLogo\":\"https://assets.ngc.nvidia.com/products/api-catalog/code-documentation-for-software-development/publisher.png\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2qNXgLMuzlJ8LS8Kl30EXbGwxSh\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/crewAIInc/nvidia-demo\"}},\"artifactName\":\"code-documentation-for-software-development\"},\"config\":{\"name\":\"code-documentation-for-software-development\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"structured-report-generation\",\"displayName\":\"Structured Report Generation\",\"publisher\":\"langchain\",\"shortDescription\":\"Generate detailed, structured reports on any topic using LangGraph and Llama3.3 70B NIM\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/structured-report-generation.jpg\",\"labels\":[\"AI Agents\",\"Blueprint\",\"LangGraph\",\"Partner\",\"Report Generation\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"},{\"key\":\"PUBLISHERLOGO\",\"value\":\"https://assets.ngc.nvidia.com/products/api-catalog/structured-report-generation/publisher.png\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:26:15.042Z\",\"description\":\"$6e\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:27:30.992Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"eb762cdd-d7cf-4ceb-90a3-613bf721b8ce\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:27:31.374Z\",\"nvcfFunctionId\":\"3c5f0b75-3c43-455f-bc94-58ff8cbb5388\",\"createdDate\":\"2025-01-07T04:26:15.364Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"publisherLogo\":\"https://assets.ngc.nvidia.com/products/api-catalog/structured-report-generation/publisher.png\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2qzwjeLSpDcjU68lNQyDcZwxmn9\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/langchain-ai/langchain-nvidia/blob/main/cookbook/structured_report_generation.ipynb\"}},\"artifactName\":\"structured-report-generation\"},\"config\":{\"name\":\"structured-report-generation\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"document-research-assistant-for-blog-creation\",\"displayName\":\"Document Research Assistant for Blog Creation\",\"publisher\":\"llamaindex\",\"shortDescription\":\"Automate research, and generate blogs with AI Agents using LlamaIndex and Llama3.3-70B NIM LLM.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/document-research-assistant-for-blog-creation.jpg\",\"labels\":[\"AI Agents\",\"Blog Creation\",\"Blueprint\",\"LlamaIndex\",\"Partner\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"},{\"key\":\"PUBLISHERLOGO\",\"value\":\"https://assets.ngc.nvidia.com/products/api-catalog/document-research-assistant-for-blog-creation/publisher.png\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:29:06.689Z\",\"description\":\"$6f\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:26:49.555Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"db659e9f-da85-48b8-aba0-1488cf396b00\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-07T18:26:49.934Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-01-07T04:29:06.923Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"publisherLogo\":\"https://assets.ngc.nvidia.com/products/api-catalog/document-research-assistant-for-blog-creation/publisher.png\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2qRrBuBdUGzzauhql87XCd7U1wR\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/run-llama/llama_index/blob/main/docs/docs/examples/agent/nvidia_document_research_assistant_for_blog_creation.ipynb\"}},\"artifactName\":\"document-research-assistant-for-blog-creation\"},\"config\":{\"name\":\"document-research-assistant-for-blog-creation\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"voice-agent-framework-for-conversational-ai\",\"displayName\":\"Voice Agent Framework for Conversational AI\",\"publisher\":\"pipecat\",\"shortDescription\":\"Automate voice AI agents with NVIDIA NIM microservices and Pipecat.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/voice-agent-framework-for-conversational-ai.jpg\",\"labels\":[\"AI Agents\",\"Blueprint\",\"Conversational AI\",\"Partner\",\"Pipecat\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"},{\"key\":\"PUBLISHERLOGO\",\"value\":\"https://assets.ngc.nvidia.com/products/api-catalog/voice-agent-framework-for-conversational-ai/publisher.png\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:32:37.900Z\",\"description\":\"$70\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-01-07T04:34:25.293Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"152727e3-00b1-4754-b892-404617836758\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-01-07T04:34:25.795Z\",\"nvcfFunctionId\":\"None\",\"createdDate\":\"2025-01-07T04:32:38.112Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"publisherLogo\":\"https://assets.ngc.nvidia.com/products/api-catalog/voice-agent-framework-for-conversational-ai/publisher.png\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy/now?launchableID=env-2qzseoG3oK2OGPQVEmrphZp5UnL\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/daily-co/nimble-pipecat\"}},\"artifactName\":\"voice-agent-framework-for-conversational-ai\"},\"config\":{\"name\":\"voice-agent-framework-for-conversational-ai\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"traceability-for-agentic-ai\",\"displayName\":\"Traceability for Agentic AI\",\"publisher\":\"wandb\",\"shortDescription\":\"Trace and evaluate AI Agents with Weights \u0026 Biases.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/traceability-for-agentic-ai.jpg\",\"labels\":[\"AI Agents\",\"Blueprint\",\"Partner\",\"Traceability\",\"WandB\",\"NVIDIA AI\"],\"attributes\":[{\"key\":\"ENTERPRISEREADY\",\"value\":\"false\"},{\"key\":\"LAUNCHABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"false\"},{\"key\":\"PUBLISHERLOGO\",\"value\":\"https://assets.ngc.nvidia.com/products/api-catalog/traceability-for-agentic-ai/publisher.png\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-07T04:35:58.243Z\",\"description\":\"$71\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-01-07T04:35:58.243Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"54d036d0-0091-4fe4-8a2b-cdee419b6041\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"nvcfFunctionId\":\"6309dd48-9048-45bc-8533-be23707890c8\",\"createdDate\":\"2025-01-07T04:35:58.499Z\",\"attributes\":{\"apiDocsUrl\":\"NOT REQUIRED\",\"publisherLogo\":\"https://assets.ngc.nvidia.com/products/api-catalog/traceability-for-agentic-ai/publisher.png\",\"termsOfUse\":\"GOVERNING TERMS: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Nemo Text Retriever E5 Embedding Model, MIT License; NVIDIA Retrieval QA Mistral 4B Reranking v3, Apache License - The use of these models is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eAI Foundation Models Community License Agreement\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement, Built with Llama\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Deploy Launchable\",\"url\":\"https://console.brev.dev/launchable/deploy?launchableID=env-2quoRI6MKlyaWI0mIQ8tVmZTOzx\"},\"secondaryCta\":{\"text\":\"View Source Code\",\"url\":\"https://github.com/wandb/ai-virtual-assistant/blob/main/deploy/ai_virtual_assistant_notebook.ipynb\"}},\"artifactName\":\"traceability-for-agentic-ai\"},\"config\":{\"name\":\"traceability-for-agentic-ai\",\"type\":\"model\"}}],\"items\":[\"$72\",\"$73\",\"$74\",\"$75\",\"$76\",\"$77\",\"$78\",\"$79\",\"$7a\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"Create AI Agents\"}],[\"$\",\"$L26\",null,{\"href\":\"/blueprints?filters=blueprintType%3Ablueprinttype_nvidia_ai\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"Blueprints to build and deploy Agentic AI applications, digital twins, etc.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"7b:T2ef0,"])</script><script>self.__next_f.push([1,"# Llama-3.3-Nemotron-Super-49B-v1\n\n\n## Model Overview \n\nLlama-3.3-Nemotron-Super-49B-v1 is a large language model (LLM) which is a derivative of Meta Llama-3.3-70B-Instruct (AKA the reference model). It is a reasoning model that is post trained for reasoning, human chat preferences, and tasks, such as RAG and tool calling. The model supports a context length of 128K tokens.\n\nLlama-3.3-Nemotron-Super-49B-v1 is a model which offers a great tradeoff between model accuracy and efficiency. Efficiency (throughput) directly translates to savings. Using a novel Neural Architecture Search (NAS) approach, we greatly reduce the model’s memory footprint, enabling larger workloads, as well as fitting the model on a single GPU at high workloads (H200). This NAS approach enables the selection of a desired point in the accuracy-efficiency tradeoff.\n\nThe model underwent a multi-phase post-training process to enhance both its reasoning and non-reasoning capabilities. This includes a supervised fine-tuning stage for Math, Code, Reasoning, and Tool Calling as well as multiple reinforcement learning (RL) stages using REINFORCE (RLOO) and Online Reward-aware Preference Optimization (RPO) algorithms for both chat and instruction-following. The final model checkpoint is obtained after merging the final SFT and Online RPO checkpoints. For more details on how the model was trained, please see [this blog](https://developer.nvidia.com/blog/build-enterprise-ai-agents-with-advanced-open-nvidia-llama-nemotron-reasoning-models/).\n\n![Architecture Diagram](https://assets.ngc.nvidia.com/products/api-catalog/llama-3_3-nemotron-super-49b-v1/diagram.jpg)\n\nThis model is part of the Llama Nemotron Collection. You can find the other model(s) in this family here: \n[Llama-3_1-Nemotron-Nano-8B-v1](https://build.nvidia.com/nvidia/llama-3_1-nemotron-nano-8b-v1)\n\nThis model is ready for commercial use. \n\n## License/Terms of Use\n\nGOVERNING TERMS: Your use of this model is governed by the [NVIDIA Open Model License.](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/) Additional Information: [Llama 3.3 Community License Agreement](https://www.llama.com/llama3_3/license/). Built with Llama.\n\n**Model Developer:** NVIDIA\n\n**Model Dates:** Trained between November 2024 and February 2025\n\n**Data Freshness:** The pretraining data has a cutoff of 2023 per Meta Llama 3.3 70B\n\n### Use Case: \u003cbr\u003e\nDevelopers designing AI Agent systems, chatbots, RAG systems, and other AI-powered applications. Also suitable for typical instruction-following tasks. \u003cbr\u003e\n\n### Release Date: \u003cbr\u003e\n3/18/2025 \u003cbr\u003e\n\n## References\n* [2502.00203] [Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment](https://arxiv.org/abs/2502.00203)\n\n## Model Architecture\n**Architecture Type:** Dense decoder-only Transformer model \n**Network Architecture:** Llama 3.3 70B Instruct, customized through Neural Architecture Search (NAS)\n\nThe model is a derivative of Meta’s Llama-3.3-70B-Instruct, using Neural Architecture Search (NAS). The NAS algorithm results in non-standard and non-repetitive blocks. This includes the following: \nSkip attention: In some blocks, the attention is skipped entirely, or replaced with a single linear layer.\nVariable FFN: The expansion/compression ratio in the FFN layer is different between blocks. \n\nWe utilize a block-wise distillation of the reference model, where for each block we create multiple variants providing different tradeoffs of quality vs. computational complexity, discussed in more depth below. We then search over the blocks to create a model which meets the required throughput and memory (optimized for a single H100-80GB GPU) while minimizing the quality degradation. The model then undergoes knowledge distillation (KD), with a focus on English single and multi-turn chat use-cases. The KD step included 40 billion tokens consisting of a mixture of 3 datasets - FineWeb, Buzz-V1.2 and Dolma.\n\n## Intended use\n\nLlama-3.3-Nemotron-Super-49B-v1 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai) are also supported. \n\n## Input\n- **Input Type:** Text\n- **Input Format:** String\n- **Input Parameters:** One-Dimensional (1D)\n- **Other Properties Related to Input:** Context length up to 131,072 tokens\n\n## Output\n- **Output Type:** Text\n- **Output Format:** String\n- **Output Parameters:** One-Dimensional (1D)\n- **Other Properties Related to Output:** Context length up to 131,072 tokens\n\n## Model Version\n1.0 (3/18/2025)\n\n## Software Integration\n- **Runtime Engine:** Transformers\n- **Recommended Hardware Microarchitecture Compatibility:** \n - NVIDIA Hopper\n - NVIDIA Ampere\n\n## Quick Start and Usage Recommendations:\n\n1. Reasoning mode (ON/OFF) is controlled via the system prompt, which must be set as shown in the example below. All instructions should be contained within the user prompt\n2. We recommend setting temperature to `0.6`, and Top P to `0.95` for Reasoning ON mode\n3. We recommend using greedy decoding for Reasoning OFF mode\n4. We have provided a list of prompts to use for evaluation for each benchmark where a specific template is required\n\nYou can try this model out through the preview API, using this link: [Llama-3_3-Nemotron-Super-49B-v1](https://build.nvidia.com/nvidia/llama-3_3-nemotron-super-49b-v1).\n\n## Inference:\n**Engine:**\nTransformers \n**Test Hardware:**\n- FP8: 1x NVIDIA H100-80GB GPU (Coming Soon!)\n- BF16: \n - 2x NVIDIA H100-80GB GPUs\n - 2x NVIDIA A100-80GB GPUs\n \n**[Preferred/Supported] Operating System(s):** Linux \u003cbr\u003e\n\n## Training Datasets\n\nA large variety of training data was used for the knowledge distillation phase before post-training pipeline, 3 of which included: FineWeb, Buzz-V1.2, and Dolma.\n\nThe data for the multi-stage post-training phases for improvements in Code, Math, and Reasoning is a compilation of SFT and RL data that supports improvements of math, code, general reasoning, and instruction following capabilities of the original Llama instruct model. \n\nIn conjunction with this model release, NVIDIA has released 30M samples of post-training data, as public and permissive. [Llama-Nemotron-Post-Training-Dataset-v1](https://huggingface.co/datasets/nvidia/Llama-Nemotron-Post-Training-Dataset-v1)\n\nDistribution of the domains is as follows:\n\n| Category | Value |\n|----------|-----------|\n| math | 19,840,970|\n| code | 9,612,677 |\n| science | 708,920 |\n| instruction following | 56,339 |\n| chat | 39,792 |\n| safety | 31,426 |\n\nPrompts have been sourced from either public and open corpus or synthetically generated. Responses were synthetically generated by a variety of models, with some prompts containing responses for both reasoning on and off modes, to train the model to distinguish between two modes. \n\nModels that were used in the creation of this dataset:\n- Llama-3.3-70B-Instruct\n- Llama-3.1-Nemotron-70B-Instruct\n- Llama-3.3-Nemotron-70B-Feedback/Edit/Select\n- Mixtral-8x22B-Instruct-v0.1\n- DeepSeek-R1\n- Qwen-2.5-Math-7B-Instruct\n- Qwen-2.5-Coder-32B-Instruct\n- Qwen-2.5-72B-Instruct\n- Qwen-2.5-32B-Instruct\n\n**Data Collection for Training Datasets:**\nHybrid: Automated, Human, Synthetic\n\n**Data Labeling for Training Datasets:**\nHybrid: Automated, Human, Synthetic\n\n## Evaluation Datasets \n\nWe used the datasets listed below to evaluate Llama-3.3-Nemotron-Super-49B-v1. \n\n**Data Collection for Evaluation Datasets:**\nHybrid: Human/Synthetic\n\n**Data Labeling for Evaluation Datasets:**\nHybrid: Human/Synthetic/Automatic\n\n## Evaluation Results\nThese results contain both Reasoning On, and Reasoning Off. We recommend using temperature=`0.6`, top_p=`0.95` for Reasoning On mode, and greedy decoding for Reasoning Off mode. All evaluations are done with 32k sequence length. We run the benchmarks up to 16 times and average the scores to be more accurate.\n\n\u003e NOTE: Where applicable, a Prompt Template will be provided. While completing benchmarks, please ensure that you are parsing for the correct output format as per the provided prompt in order to reproduce the benchmarks seen below. \n\n### Arena-Hard\n\n| Reasoning Mode | Score |\n|--------------|------------|\n| Reasoning Off | 88.3 | \n\n\n### MATH500\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 74.0 | \n| Reasoning On | 96.6 |\n\nUser Prompt Template: \n```\n\"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \\boxed{}.\\nQuestion: {question}\"\n```\n\n### AIME25\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 13.33 | \n| Reasoning On | 58.4 |\n\nUser Prompt Template: \n```\n\"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \\boxed{}.\\nQuestion: {question}\"\n```\n\n### GPQA\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 50 | \n| Reasoning On | 66.67 |\n\nUser Prompt Template: \n```\n\"What is the correct answer to this question: {question}\\nChoices:\\nA. {option_A}\\nB. {option_B}\\nC. {option_C}\\nD. {option_D}\\nLet's think step by step, and put the final answer (should be a single letter A, B, C, or D) into a \\boxed{}\"\n```\n\n### IFEval\n\n| Reasoning Mode | Strict:Instruction |\n|--------------|------------|\n| Reasoning Off | 89.21 | \n\n### BFCL V2 Live\n\n| Reasoning Mode | Score |\n|--------------|------------|\n| Reasoning Off | 73.7 | \n\nUser Prompt Template:\n```\nYou are an expert in composing functions. You are given a question and a set of possible functions. \nBased on the question, you will need to make one or more function/tool calls to achieve the purpose. \nIf none of the function can be used, point it out. If the given question lacks the parameters required by the function,\nalso point it out. You should only return the function call in tools call sections.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of \u003cTOOLCALL\u003e[func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\u003c/TOOLCALL\u003e\n\nYou SHOULD NOT include any other text in the response.\nHere is a list of functions in JSON format that you can invoke.\n\n\u003cAVAILABLE_TOOLS\u003e{functions}\u003c/AVAILABLE_TOOLS\u003e\n\n{user_prompt}\n```\n\n### MBPP 0-shot\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 84.9| \n| Reasoning On | 91.3 |\n\nUser Prompt Template:\n````\nYou are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.\n\n@@ Instruction\nHere is the given problem and test examples:\n{prompt}\nPlease use the python programming language to solve this problem.\nPlease make sure that your code includes the functions from the test samples and that the input and output formats of these functions match the test samples.\nPlease return all completed codes in one code block.\nThis code block should be in the following format:\n```python\n# Your codes here\n```\n````\n\n### MT-Bench\n\n| Reasoning Mode | Score |\n|--------------|------------|\n| Reasoning Off | 9.17 |\n\n## Ethical Considerations:\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. \n\nFor more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards, which you can find by clicking the ModelCard++ tab above, next to Overview.\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"7c:Tb66,"])</script><script>self.__next_f.push([1,"|Field:|Response:|\n|:---:|:---:|\n|Intended Application(s) \u0026 Domain(s):| Text generation, reasoning, summarization, and question answering. |\n|Model Type: |Text-to-text transformer |\n|Intended Users:|This model is intended for developers, researchers, and customers building/utilizing LLMs, while balancing accuracy and efficiency.|\n|Output:|Text String(s)|\n|Describe how the model works:|Generates text by predicting the next word or token based on the context provided in the input sequence using multiple self-attention layers|\n|Technical Limitations:| The model was trained on data that contains toxic language, unsafe content, and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.\u003cbr/\u003eThe model demonstrates weakness to alignment-breaking attacks. Users are advised to deploy language model guardrails alongside this model to prevent potentially harmful outputs.\u003cbr/\u003eThe Model may generate answers that are inaccurate, omit key information, or include irrelevant or redundant text.|\n|Verified to have met prescribed quality standards?|Yes|\n|Performance Metrics:|Accuracy, Throughput, and user-side throughput|\n|Potential Known Risks:|The model was optimized explicitly for instruction following and as such is more susceptible to prompt injection and jailbreaking in various forms as a result of its instruction tuning. This means that the model should be paired with additional rails or system filtering to limit exposure to instructions from malicious sources -- either directly or indirectly by retrieval (e.g. via visiting a website) -- as they may yield outputs that can lead to harmful, system-level outcomes up to and including remote code execution in agentic systems when effective security controls including guardrails are not in place.\u003cbr/\u003eThe model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.|\n|End User License Agreement:| Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.3 Community License Agreement](https://www.llama.com/llama3_3/license/). Built with Llama. |"])</script><script>self.__next_f.push([1,"7d:T4eb,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/llama-3.3-nemotron-super-49b-v1:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"nvidia/llama-3.3-nemotron-super-49b-v1\",\n \"messages\": [{\"role\": \"system\", \"content\": \"detailed thinking off\"}, {\"role\":\"user\", \"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).7e:T138d,"])</script><script>self.__next_f.push([1,"**Model Overview**\n\n## Description:\nDeepSeek-R1 is a first-generation reasoning model trained using large-scale reinforcement learning (RL) to solve complex reasoning tasks across domains such as math, code, and language. The model leverages RL to develop reasoning capabilities, which are further enhanced through supervised fine-tuning (SFT) to improve readability and coherence. DeepSeek-R1 achieves state-of-the-art results in various benchmarks and offers both its base models and distilled versions for community use.\n\nThis model is ready for both research and commercial use.\nFor more details, visit the [DeepSeek website](https://www.deepseek.com/).\n\n\u003cdiv align=\"center\"\u003e\n\u003cimg src=\"https://huggingface.co/deepseek-ai/DeepSeek-R1/resolve/main/figures/benchmark.jpg\" width=\"80%\" alt=\"Benchmarking\"/\u003e\n\u003c/div\u003e\n\n## Third-Party Community Consideration:\nThis model is not owned or developed by NVIDIA. It is a community-driven model created by DeepSeek AI. See the official [DeepSeek-R1 Model Card](https://huggingface.co/deepseek-ai/DeepSeek-R1) on Hugging Face for further details.\n\n## License/Terms of Use:\nGOVERNING TERMS: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). Additional Information: [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md).\n\n## References:\n- [DeepSeek GitHub Repository](https://github.com/deepseek-ai/DeepSeek-R1)\n- [DeepSeek-R1 Paper](https://arxiv.org/abs/2501.12948)\n\n## Model Architecture:\n\n**Architecture Type:** Mixture of Experts (MoE) \u003cbr\u003e\n**Network Architecture:** \u003cbr\u003e\n- Base Model: DeepSeek-V3-Base\n- Activated Parameters: 37 billion\n- Total Parameters: 671 billion\n- Distilled Models: Smaller, fine-tuned versions based on Qwen and Llama architectures.\n- Context Length: 128K tokens\n\n## Input:\n\n**Input Type(s):** Text \u003cbr\u003e\n**Input Format(s):** String \u003cbr\u003e\n**Input Parameters:** (1D) \u003cbr\u003e\n**Other Properties Related to Input:** \u003cbr\u003e\nDeepSeek recommends adhering to the following configurations when utilizing the DeepSeek-R1 series models, including benchmarking, to achieve the expected performance:\n\n1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.\n2. **Avoid adding a system prompt; all instructions should be contained within the user prompt.**\n3. For mathematical problems, it is advisable to include a directive in your prompt such as: \"Please reason step by step, and put your final answer within \\boxed{}.\"\n4. When evaluating model performance, it is recommended to conduct multiple tests and average the results.\n\n## Output:\n**Output Type(s):** Text \u003cbr\u003e\n**Output Format:** String \u003cbr\u003e\n**Output Parameters:** (1D) \u003cbr\u003e\n\n## Software Integration:\n**Runtime Engine(s):** vLLM and SGLang \u003cbr\u003e\n**Supported Hardware Microarchitecture Compatibility:** NVIDIA Ampere, NVIDIA Blackwell, NVIDIA Jetson, NVIDIA Hopper, NVIDIA Lovelace, NVIDIA Pascal, NVIDIA Turing, and NVIDIA Volta architectures \u003cbr\u003e\n**[Preferred/Supported] Operating System(s):** Linux\n\n## Model Version(s):\nDeepSeek-R1 V1.0\n\n## Training, Testing, and Evaluation Datasets:\n### Training Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n### Testing Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n### Evaluation Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n## Inference:\n**Engine:** SGLang\n**Test Hardware:** NVIDIA Hopper\n\n## Ethical Considerations:\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Model Limitations:\nThe base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive."])</script><script>self.__next_f.push([1,"7f:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\n80:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n "])</script><script>self.__next_f.push([1," messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();81:T667,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choic"])</script><script>self.__next_f.push([1,"e).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003e82:T485,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/deepseek-ai/deepseek-r1:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"deepseek-ai/deepseek-r1\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Which number is larger, 9.11 or 9.8?\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).83:T230d,"])</script><script>self.__next_f.push([1,"# Llama-3.1-Nemotron-Nano-8B-v1\n\n\n## Model Overview \n\nLlama-3.1-Nemotron-Nano-8B-v1 is a large language model (LLM) which is a derivative of Meta Llama-3.1-8B-Instruct (AKA the reference model). It is a reasoning model that is post trained for reasoning, human chat preferences, and tasks, such as RAG and tool calling. \n\nLlama-3.1-Nemotron-Nano-8B-v1 is a model which offers a great tradeoff between model accuracy and efficiency. It is created from Llama 3.1 8B Instruct and offers improvements in model accuracy. The model fits on a single RTX GPU and can be used locally. The model supports a context length of 128K.\n\nThis model underwent a multi-phase post-training process to enhance both its reasoning and non-reasoning capabilities. This includes a supervised fine-tuning stage for Math, Code, Reasoning, and Tool Calling as well as multiple reinforcement learning (RL) stages using REINFORCE (RLOO) and Online Reward-aware Preference Optimization (RPO) algorithms for both chat and instruction-following. The final model checkpoint is obtained after merging the final SFT and Online RPO checkpoints.\n\nThis model is part of the Llama Nemotron Collection. You can find the other model(s) in this family here: \n[Llama-3_3-Nemotron-Super-49B-v1](https://build.nvidia.com/nvidia/llama-3_3-nemotron-super-49b-v1)\n\nThis model is ready for commercial use.\n\n## License/Terms of Use\n\nGOVERNING TERMS: Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.1 Community License Agreement](https://www.llama.com/llama3_1/license/). Built with Llama.\n\n**Model Developer:** NVIDIA\n\n**Model Dates:** Trained between August 2024 and March 2025\n\n**Data Freshness:** The pretraining data has a cutoff of 2023 per Meta Llama 3.1 8B\n\n\n## Use Case: \n\nDevelopers designing AI Agent systems, chatbots, RAG systems, and other AI-powered applications. Also suitable for typical instruction-following tasks. Balance of model accuracy and compute efficiency (the model fits on a single RTX GPU and can be used locally).\n\n## Release Date: \u003cbr\u003e\n3/18/2025 \u003cbr\u003e\n\n## References\n* [2502.00203] [Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment](https://arxiv.org/abs/2502.00203)\n\n\n## Model Architecture\n**Architecture Type:** Dense decoder-only Transformer model \n**Network Architecture:** Llama 3.1 8B Instruct\n\n## Intended use\n\nLlama-3.1-Nemotron-Nano-8B-v1 is a general purpose reasoning and chat model intended to be used in English and coding languages. Other non-English languages (German, French, Italian, Portuguese, Hindi, Spanish, and Thai) are also supported. \n\n## Input:\n- **Input Type:** Text\n- **Input Format:** String\n- **Input Parameters:** One-Dimensional (1D)\n- **Other Properties Related to Input:** Context length up to 131,072 tokens\n\n## Output:\n- **Output Type:** Text\n- **Output Format:** String\n- **Output Parameters:** One-Dimensional (1D)\n- **Other Properties Related to Output:** Context length up to 131,072 tokens\n\n## Model Version:\n1.0 (3/18/2025)\n\n## Software Integration\n- **Runtime Engine:** NeMo 24.12 \u003cbr\u003e\n- **Recommended Hardware Microarchitecture Compatibility:**\n - NVIDIA Hopper\n - NVIDIA Ampere\n\n## Quick Start and Usage Recommendations:\n\n1. Reasoning mode (ON/OFF) is controlled via the system prompt, which must be set as shown in the example below. All instructions should be contained within the user prompt\n2. We recommend setting temperature to `0.6`, and Top P to `0.95` for Reasoning ON mode\n3. We recommend using greedy decoding for Reasoning OFF mode\n4. We have provided a list of prompts to use for evaluation for each benchmark where a specific template is required\n\nYou can try this model out through the preview API, using this link: [Llama-3_1-Nemotron-Nano-8B-v1](https://build.nvidia.com/nvidia/llama-3_1-nemotron-nano-8b-v1).\n\n## Inference:\n**Engine:** Transformers \n**Test Hardware:**\n- BF16:\n - 1x RTX 50 Series GPUs\n - 1x RTX 40 Series GPUs\n - 1x RTX 30 Series GPUs\n - 1x H100-80GB GPU\n - 1x A100-80GB GPU\n\n\n\n**Preferred/Supported] Operating System(s):** Linux \u003cbr\u003e\n\n## Training Datasets\n\nA large variety of training data was used for the post-training pipeline, including manually annotated data and synthetic data.\n\nThe data for the multi-stage post-training phases for improvements in Code, Math, and Reasoning is a compilation of SFT and RL data that supports improvements of math, code, general reasoning, and instruction following capabilities of the original Llama instruct model. \n\nPrompts have been sourced from either public and open corpus or synthetically generated. Responses were synthetically generated by a variety of models, with some prompts containing responses for both Reasoning On and Off modes, to train the model to distinguish between two modes. \n\n**Data Collection for Training Datasets:** Hybrid: Automated, Human, Synthetic \u003cbr\u003e\n\n**Data Labeling for Training Datasets:** N/A \u003cbr\u003e\n\n## Evaluation Datasets\n\nWe used the datasets listed below to evaluate Llama-3.1-Nemotron-Nano-8B-v1. \n\n**Data Collection for Evaluation Datasets:** Hybrid: Human/Synthetic\n\n**Data Labeling for Evaluation Datasets:** Hybrid: Human/Synthetic/Automatic\n\n## Evaluation Results\n\nThese results contain both “Reasoning On”, and “Reasoning Off”. We recommend using temperature=`0.6`, top_p=`0.95` for “Reasoning On” mode, and greedy decoding for “Reasoning Off” mode. All evaluations are done with 32k sequence length. We run the benchmarks up to 16 times and average the scores to be more accurate.\n\n\u003e NOTE: Where applicable, a Prompt Template will be provided. While completing benchmarks, please ensure that you are parsing for the correct output format as per the provided prompt in order to reproduce the benchmarks seen below. \n\n### MT-Bench\n\n| Reasoning Mode | Score |\n|--------------|------------|\n| Reasoning Off | 7.9 |\n| Reasoning On | 8.1 |\n\n\n### MATH500\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 36.6% | \n| Reasoning On | 95.4% |\n\nUser Prompt Template: \n```\n\"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \\boxed{}.\\nQuestion: {question}\"\n```\n\n### AIME25\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 36.6% | \n| Reasoning On | 47.1% |\n\nUser Prompt Template: \n```\n\"Below is a math question. I want you to reason through the steps and then give a final answer. Your final answer should be in \\boxed{}.\\nQuestion: {question}\"\n```\n\n### GPQA-D\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 39.4% | \n| Reasoning On | 54.1% |\n\nUser Prompt Template: \n```\n\"What is the correct answer to this question: {question}\\nChoices:\\nA. {option_A}\\nB. {option_B}\\nC. {option_C}\\nD. {option_D}\\nLet's think step by step, and put the final answer (should be a single letter A, B, C, or D) into a \\boxed{}\"\n```\n\n### IFEval Average\n\n| Reasoning Mode | Strict:Prompt | Strict:Instruction |\n|--------------|------------|------------|\n| Reasoning Off | 74.7% | 82.1% |\n| Reasoning On | 71.9% | 79.3% |\n\n### BFCL v2 Live\n\n| Reasoning Mode | Score |\n|--------------|------------|\n| Reasoning Off | 63.9% | \n| Reasoning On | 63.6% | \n\nUser Prompt Template:\n```\n\u003cAVAILABLE_TOOLS\u003e{functions}\u003c/AVAILABLE_TOOLS\u003e\n\n{user_prompt}\n```\n\n### MBPP 0-shot\n\n| Reasoning Mode | pass@1 |\n|--------------|------------|\n| Reasoning Off | 66.1% | \n| Reasoning On | 84.6% |\n\nUser Prompt Template:\n````\nYou are an exceptionally intelligent coding assistant that consistently delivers accurate and reliable responses to user instructions.\n\n@@ Instruction\nHere is the given problem and test examples:\n{prompt}\nPlease use the python programming language to solve this problem.\nPlease make sure that your code includes the functions from the test samples and that the input and output formats of these functions match the test samples.\nPlease return all completed codes in one code block.\nThis code block should be in the following format:\n```python\n# Your codes here\n```\n````\n\n## Ethical Considerations:\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. \n\nFor more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards, which you can find by clicking the ModelCard++ tab above, next to Overview.\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"84:Tb66,"])</script><script>self.__next_f.push([1,"|Field:|Response:|\n|:---:|:---:|\n|Intended Application(s) \u0026 Domain(s):| Text generation, reasoning, summarization, and question answering. |\n|Model Type: |Text-to-text transformer |\n|Intended Users:|This model is intended for developers, researchers, and customers building/utilizing LLMs, while balancing accuracy and efficiency.|\n|Output:|Text String(s)|\n|Describe how the model works:|Generates text by predicting the next word or token based on the context provided in the input sequence using multiple self-attention layers|\n|Technical Limitations:| The model was trained on data that contains toxic language, unsafe content, and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.\u003cbr/\u003eThe model demonstrates weakness to alignment-breaking attacks. Users are advised to deploy language model guardrails alongside this model to prevent potentially harmful outputs.\u003cbr/\u003eThe Model may generate answers that are inaccurate, omit key information, or include irrelevant or redundant text.|\n|Verified to have met prescribed quality standards?|Yes|\n|Performance Metrics:|Accuracy, Throughput, and user-side throughput|\n|Potential Known Risks:|The model was optimized explicitly for instruction following and as such is more susceptible to prompt injection and jailbreaking in various forms as a result of its instruction tuning. This means that the model should be paired with additional rails or system filtering to limit exposure to instructions from malicious sources -- either directly or indirectly by retrieval (e.g. via visiting a website) -- as they may yield outputs that can lead to harmful, system-level outcomes up to and including remote code execution in agentic systems when effective security controls including guardrails are not in place.\u003cbr/\u003eThe model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.|\n|End User License Agreement:| Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.1 Community License Agreement](https://www.llama.com/llama3_1/license/). Built with Llama. |"])</script><script>self.__next_f.push([1,"85:T56a4,"])</script><script>self.__next_f.push([1,"# Gemma 3 model\n\n## Description\n\nGemma is a family of lightweight, state-of-the-art open models from Google,\nbuilt from the same research and technology used to create the Gemini models.\nGemma 3 models are multimodal, handling text and image input and generating text\noutput, with open weights for both pre-trained variants and instruction-tuned\nvariants. Gemma 3 has a large, 128K context window, multilingual support in over\n140 languages, and is available in more sizes than previous versions. Gemma 3\nmodels are well-suited for a variety of text generation and image understanding\ntasks, including question answering, summarization, and reasoning. Their\nrelatively small size makes it possible to deploy them in environments with\nlimited resources such as laptops, desktops or your own cloud infrastructure,\ndemocratizing access to state of the art AI models and helping foster innovation\nfor everyone. This model is ready for commercial use.\n\n## Third-Party Community Consideration\nThis model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA [Gemma 3](https://ai.google.dev/gemma/docs/core) model card.\n\n## License/Terms of Use\nGOVERNING TERMS: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); and the use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). ADDITIONAL INFORMATION: [Gemma Terms of Use](https://ai.google.dev/gemma/terms).\n\n## Deployment Geography\nGlobal \n\n## Use Case\nModels are well-suited for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning.\n\n## Benefits\n\nAt the time of release, this family of models provides high-performance open\nvision-language model implementations designed from the ground up for\nresponsible AI development compared to similarly sized models.\n\nUsing the benchmark evaluation metrics described in this document, these models\nhave shown to provide superior performance to other, comparably-sized open model\nalternatives.\n\n## Release Date\n* **Build.Nvidia.com** - 3/11/2025 via [https://build.nvidia.com/google/gemma-3-1b-it](https://build.nvidia.com/google/gemma-3-1b-it) and [https://build.nvidia.com/google/gemma-3-27b-it](https://build.nvidia.com/google/gemma-3-27b-it)\n\n## References\n**Model Page**: [Gemma](https://ai.google.dev/gemma/docs/core) \n**Authors**: Google DeepMind\n\n* [Gemma 3 Technical Report][g3-tech-report]\n* [Responsible Generative AI Toolkit][rai-toolkit]\n* [Gemma on Kaggle][kaggle-gemma]\n* [Gemma on Vertex Model Garden][vertex-mg-gemma3]\n\n## Model Architecture\n**Architecture Type**: Dense decoder-only Transformer model\n\n### Inputs and outputs\n\n### Input\n**Input Type(s)**: Text, Text+Image \n**Input Format(s)**:\n - String\n - Image: jpg\n\n**Input Parameters**:\n- **Text**: One-dimensional (1D)\n- **Image**: Two-dimensional (2D)\n\n**Other Properties Related to Input**:\n- Text string, such as a question, a prompt, or a document to be summarized\n- Images, normalized to 896 x 896 resolution and encoded to 256 tokens\n each\n- Total input context of 128K tokens for the 4B, 12B, and 27B sizes, and\n 32K tokens for the 1B size\n\n### Output\n\n**Output Type(s)**: Text \n**Output Format**: String \n**Output Parameters**: (1D) \n**Other Properties Related to Output**:\n- Generated text in response to the input, such as an answer to a\n question, analysis of image content, or a summary of a document\n- Total output context of 8192 tokens\n\n## Software Integration\n**Runtime Engine(s)**: TRT-LLM \n**Supported Hardware Microarchitecture Compatibility**: NVIDIA Ampere, NVIDIA Blackwell, NVIDIA Jetson, NVIDIA Hopper, NVIDIA Lovelace, NVIDIA Pascal, NVIDIA Turing, and NVIDIA Volta architectures \n**[Preferred/Supported] Operating System(s)**: Linux \n\n## Model Version(s):\n* **Gemma 3 IT 1B**: 1.0 (3/12/2025)\n* **Gemma 3 IT 4B**: 1.0 (3/12/2025)\n* **Gemma 3 IT 12B**: 1.0 (3/12/2025)\n* **Gemma 3 IT 27B**: 1.0 (3/12/2025)\n\n### Software\n\nTraining was done using [JAX][jax] and [ML Pathways][ml-pathways].\n\nJAX allows researchers to take advantage of the latest generation of hardware,\nincluding TPUs, for faster and more efficient training of large models. ML\nPathways is Google's latest effort to build artificially intelligent systems\ncapable of generalizing across multiple tasks. This is especially suitable for\nfoundation models, including large language models like these ones.\n\nTogether, JAX and ML Pathways are used as described in the\n[paper about the Gemini family of models][gemini-2-paper]; *\"the 'single\ncontroller' programming model of Jax and Pathways allows a single Python\nprocess to orchestrate the entire training run, dramatically simplifying the\ndevelopment workflow.\"*\n\n## Training, Testing, and Evaluation Datasets\n### Training Dataset\n**Data Collection Method by dataset**: Hybrid: Human, Automated \n**Labeling Method by dataset**: Hybrid: Human, Automated\n\nThese models were trained on a dataset of text data that includes a wide variety\nof sources. The 27B model was trained with 14 trillion tokens, the 12B model was\ntrained with 12 trillion tokens, 4B model was trained with 4 trillion tokens and\n1B with 2 trillion tokens. Here are the key components:\n\n- Web Documents: A diverse collection of web text ensures the model is\n exposed to a broad range of linguistic styles, topics, and vocabulary. The\n training dataset includes content in over 140 languages.\n- Code: Exposing the model to code helps it to learn the syntax and\n patterns of programming languages, which improves its ability to generate\n code and understand code-related questions.\n- Mathematics: Training on mathematical text helps the model learn logical\n reasoning, symbolic representation, and address mathematical queries.\n- Images: A wide range of images enables the model to perform image\n analysis and visual data extraction tasks.\n\nThe combination of these diverse data sources is crucial for training a powerful\nmultimodal model that can handle a wide variety of different tasks and data\nformats.\n\n#### Data Preprocessing\n\nHere are the key data cleaning and filtering methods applied to the training\ndata:\n\n- CSAM Filtering: Rigorous CSAM (Child Sexual Abuse Material) filtering\n was applied at multiple stages in the data preparation process to ensure\n the exclusion of harmful and illegal content.\n- Sensitive Data Filtering: As part of making Gemma pre-trained models\n safe and reliable, automated techniques were used to filter out certain\n personal information and other sensitive data from training sets.\n- Additional methods: Filtering based on content quality and safety in\n line with [Google Responsible AI policies][safety-policies].\n\n### Testing Dataset\n**Data Collection Method by dataset**: Hybrid: Human, Automated \n**Labeling Method by dataset**: Hybrid: Human, Automated\n\n### Evaluation Dataset\n**Data Collection Method by dataset**: Hybrid: Human, Automated \n**Labeling Method by dataset**: Hybrid: Human, Automated\n\n## Evaluation\n\nModel evaluation metrics and results are highlighted below.\n\n### Benchmark Results\n\nThese models were evaluated against a large collection of different datasets and\nmetrics to cover different aspects of text generation:\n\n#### Reasoning and factuality\n\n| Benchmark | Metric | Gemma 3 PT 1B | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |\n| ------------------------------ |----------------|:--------------:|:-------------:|:--------------:|:--------------:|\n| [HellaSwag][hellaswag] | 10-shot | 62.3 | 77.2 | 84.2 | 85.6 |\n| [BoolQ][boolq] | 0-shot | 63.2 | 72.3 | 78.8 | 82.4 |\n| [PIQA][piqa] | 0-shot | 73.8 | 79.6 | 81.8 | 83.3 |\n| [SocialIQA][socialiqa] | 0-shot | 48.9 | 51.9 | 53.4 | 54.9 |\n| [TriviaQA][triviaqa] | 5-shot | 39.8 | 65.8 | 78.2 | 85.5 |\n| [Natural Questions][naturalq] | 5-shot | 9.48 | 20.0 | 31.4 | 36.1 |\n| [ARC-c][arc] | 25-shot | 38.4 | 56.2 | 68.9 | 70.6 |\n| [ARC-e][arc] | 0-shot | 73.0 | 82.4 | 88.3 | 89.0 |\n| [WinoGrande][winogrande] | 5-shot | 58.2 | 64.7 | 74.3 | 78.8 |\n| [BIG-Bench Hard][bbh] | few-shot | 28.4 | 50.9 | 72.6 | 77.7 |\n| [DROP][drop] | 1-shot | 42.4 | 60.1 | 72.2 | 77.2 |\n\n[hellaswag]: https://arxiv.org/abs/1905.07830\n[boolq]: https://arxiv.org/abs/1905.10044\n[piqa]: https://arxiv.org/abs/1911.11641\n[socialiqa]: https://arxiv.org/abs/1904.09728\n[triviaqa]: https://arxiv.org/abs/1705.03551\n[naturalq]: https://github.com/google-research-datasets/natural-questions\n[arc]: https://arxiv.org/abs/1911.01547\n[winogrande]: https://arxiv.org/abs/1907.10641\n[bbh]: https://paperswithcode.com/dataset/bbh\n[drop]: https://arxiv.org/abs/1903.00161\n\n#### STEM and code\n\n| Benchmark | Metric | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |\n| ------------------------------ |----------------|:-------------:|:--------------:|:--------------:|\n| [MMLU][mmlu] | 5-shot | 59.6 | 74.5 | 78.6 |\n| [MMLU][mmlu] (Pro COT) | 5-shot | 29.2 | 45.3 | 52.2 |\n| [AGIEval][agieval] | 3-5-shot | 42.1 | 57.4 | 66.2 |\n| [MATH][math] | 4-shot | 24.2 | 43.3 | 50.0 |\n| [GSM8K][gsm8k] | 8-shot | 38.4 | 71.0 | 82.6 |\n| [GPQA][gpqa] | 5-shot | 15.0 | 25.4 | 24.3 |\n| [MBPP][mbpp] | 3-shot | 46.0 | 60.4 | 65.6 |\n| [HumanEval][humaneval] | 0-shot | 36.0 | 45.7 | 48.8 |\n\n[mmlu]: https://arxiv.org/abs/2009.03300\n[agieval]: https://arxiv.org/abs/2304.06364\n[math]: https://arxiv.org/abs/2103.03874\n[gsm8k]: https://arxiv.org/abs/2110.14168\n[gpqa]: https://arxiv.org/abs/2311.12022\n[mbpp]: https://arxiv.org/abs/2108.07732\n[humaneval]: https://arxiv.org/abs/2107.03374\n\n#### Multilingual\n\n| Benchmark | Gemma 3 PT 1B | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |\n| ------------------------------------ |:-------------:|:-------------:|:--------------:|:--------------:|\n| [MGSM][mgsm] | 2.04 | 34.7 | 64.3 | 74.3 |\n| [Global-MMLU-Lite][global-mmlu-lite] | 24.9 | 57.0 | 69.4 | 75.7 |\n| [WMT24++][wmt24pp] (ChrF) | 36.7 | 48.4 | 53.9 | 55.7 |\n| [FloRes][flores] | 29.5 | 39.2 | 46.0 | 48.8 |\n| [XQuAD][xquad] (all) | 43.9 | 68.0 | 74.5 | 76.8 |\n| [ECLeKTic][eclektic] | 4.69 | 11.0 | 17.2 | 24.4 |\n| [IndicGenBench][indicgenbench] | 41.4 | 57.2 | 61.7 | 63.4 |\n\n[mgsm]: https://arxiv.org/abs/2210.03057\n[flores]: https://arxiv.org/abs/2106.03193\n[xquad]: https://arxiv.org/abs/1910.11856v3\n[global-mmlu-lite]: https://huggingface.co/datasets/CohereForAI/Global-MMLU-Lite\n[wmt24pp]: https://arxiv.org/abs/2502.12404v1\n[eclektic]: https://arxiv.org/abs/2502.21228\n[indicgenbench]: https://arxiv.org/abs/2404.16816\n\n#### Multimodal\n\n| Benchmark | Gemma 3 PT 4B | Gemma 3 PT 12B | Gemma 3 PT 27B |\n| ------------------------------ |:-------------:|:--------------:|:--------------:|\n| [COCOcap][coco-cap] | 102 | 111 | 116 |\n| [DocVQA][docvqa] (val) | 72.8 | 82.3 | 85.6 |\n| [InfoVQA][info-vqa] (val) | 44.1 | 54.8 | 59.4 |\n| [MMMU][mmmu] (pt) | 39.2 | 50.3 | 56.1 |\n| [TextVQA][textvqa] (val) | 58.9 | 66.5 | 68.6 |\n| [RealWorldQA][realworldqa] | 45.5 | 52.2 | 53.9 |\n| [ReMI][remi] | 27.3 | 38.5 | 44.8 |\n| [AI2D][ai2d] | 63.2 | 75.2 | 79.0 |\n| [ChartQA][chartqa] | 63.6 | 74.7 | 76.3 |\n| [VQAv2][vqav2] | 63.9 | 71.2 | 72.9 |\n| [BLINK][blinkvqa] | 38.0 | 35.9 | 39.6 |\n| [OKVQA][okvqa] | 51.0 | 58.7 | 60.2 |\n| [TallyQA][tallyqa] | 42.5 | 51.8 | 54.3 |\n| [SpatialSense VQA][ss-vqa] | 50.9 | 60.0 | 59.4 |\n| [CountBenchQA][countbenchqa] | 26.1 | 17.8 | 68.0 |\n\n[coco-cap]: https://cocodataset.org/#home\n[docvqa]: https://www.docvqa.org/\n[info-vqa]: https://arxiv.org/abs/2104.12756\n[mmmu]: https://arxiv.org/abs/2311.16502\n[textvqa]: https://textvqa.org/\n[realworldqa]: https://paperswithcode.com/dataset/realworldqa\n[remi]: https://arxiv.org/html/2406.09175v1\n[ai2d]: https://allenai.org/data/diagrams\n[chartqa]: https://arxiv.org/abs/2203.10244\n[vqav2]: https://visualqa.org/index.html\n[blinkvqa]: https://arxiv.org/abs/2404.12390\n[okvqa]: https://okvqa.allenai.org/\n[tallyqa]: https://arxiv.org/abs/1810.12440\n[ss-vqa]: https://arxiv.org/abs/1908.02660\n[countbenchqa]: https://github.com/google-research/big_vision/blob/main/big_vision/datasets/countbenchqa/\n\n## Inference\n**Engine**: Transformers \n**Test Hardware**: NVIDIA Hopper\n\n## Ethics and Safety\n\nEthics and safety evaluation approach and results are highlighted below.\n\n### Evaluation Approach\n\nThe evaluation method included structured evaluations and internal red-teaming\ntesting of relevant content policies. Red-teaming was conducted by a number of\ndifferent teams, each with different goals and human evaluation metrics. These\nmodels were evaluated against a number of different categories relevant to\nethics and safety, including:\n\n- **Child Safety**: Evaluation of text-to-text and image to text prompts\n covering child safety policies, including child sexual abuse and\n exploitation.\n- **Content Safety:** Evaluation of text-to-text and image to text prompts\n covering safety policies including, harassment, violence and gore, and hate\n speech.\n- **Representational Harms**: Evaluation of text-to-text and image to text\n prompts covering safety policies including bias, stereotyping, and harmful\n associations or inaccuracies.\n\nIn addition to development level evaluations, assurance evaluations were conducted using the \"arms-length\" internal evaluations for responsibility\ngovernance decision making. They are conducted separately from the model\ndevelopment team, to inform decision making about release. High level findings\nare fed back to the model team, but prompt sets are held-out to prevent\noverfitting and preserve the results' ability to inform decision making.\nAssurance evaluation results are reported to the Responsibility \u0026 Safety Council\nas part of release review.\n\n### Evaluation Results\n\nFor all areas of safety testing, there were major improvements in the categories of\nchild safety, content safety, and representational harms relative to previous\nGemma models. All testing was conducted without safety filters to evaluate the\nmodel capabilities and behaviors. For both text-to-text and image-to-text, and\nacross all model sizes, the model produced minimal policy violations, and showed\nsignificant improvements over previous Gemma models' performance with respect\nto ungrounded inferences. One limitation of the evaluation was that the models \nincorporated only English language prompts.\n\n## Usage and Limitations\n\nThe potential limitations for these models are outlined below.\n\n### Intended Usage\n\nOpen vision-language models (VLMs) models have a wide range of applications\nacross various industries and domains. The following list of potential uses is\nnot comprehensive. The purpose of this list is to provide contextual information\nabout the possible use-cases that the model creators considered as part of model\ntraining and development.\n\n- Content Creation and Communication\n - Text Generation: These models can be used to generate creative text\n formats such as poems, scripts, code, marketing copy, and email drafts.\n - Chatbots and Conversational AI: Power conversational interfaces\n for customer service, virtual assistants, or interactive applications.\n - Text Summarization: Generate concise summaries of a text corpus,\n research papers, or reports.\n - Image Data Extraction: These models can be used to extract,\n interpret, and summarize visual data for text communications.\n- Research and Education\n - Natural Language Processing (NLP) and VLM Research: These\n models can serve as a foundation for researchers to experiment with VLM\n and NLP techniques, develop algorithms, and contribute to the\n advancement of the field.\n - Language Learning Tools: Support interactive language learning\n experiences, aiding in grammar correction or providing writing practice.\n - Knowledge Exploration: Assist researchers in exploring large\n bodies of text by generating summaries or answering questions about\n specific topics.\n\n#### Model Limitations\n- Training Data\n - The quality and diversity of the training data significantly\n influence the model's capabilities. Biases or gaps in the training data\n can lead to limitations in the model's responses.\n - The scope of the training dataset determines the subject areas\n the model can handle effectively.\n- Context and Task Complexity\n - Models are better at tasks that can be framed with clear\n prompts and instructions. Open-ended or highly complex tasks might be\n challenging.\n - A model's performance can be influenced by the amount of context\n provided (longer context generally leads to better outputs, up to a\n certain point).\n- Language Ambiguity and Nuance\n - Natural language is inherently complex. Models might struggle\n to grasp subtle nuances, sarcasm, or figurative language.\n- Factual Accuracy\n - Models generate responses based on information they learned\n from their training datasets, but they are not knowledge bases. They\n may generate incorrect or outdated factual statements.\n- Common Sense\n - Models rely on statistical patterns in language. They might\n lack the ability to apply common sense reasoning in certain situations. \n\n#### Identified risks and mitigations\n\n- **Perpetuation of biases**: It's encouraged to perform continuous\n monitoring (using evaluation metrics, human review) and the exploration of\n de-biasing techniques during model training, fine-tuning, and other use\n cases.\n- **Generation of harmful content**: Mechanisms and guidelines for content\n safety are essential. Developers are encouraged to exercise caution and\n implement appropriate content safety safeguards based on their specific\n product policies and application use cases.\n- **Misuse for malicious purposes**: Technical limitations and developer\n and end-user education can help mitigate against malicious applications of\n VLMs. Educational resources and reporting mechanisms for users to flag\n misuse are provided. Prohibited uses of Gemma models are outlined in the\n [Gemma Prohibited Use Policy][prohibited-use].\n- **Privacy violations**: Models were trained on data filtered for removal\n of certain personal information and other sensitive data. Developers are\n encouraged to adhere to privacy regulations with privacy-preserving\n techniques.\n\n[g3-tech-report]: https://goo.gle/Gemma3Report\n[rai-toolkit]: https://ai.google.dev/responsible\n[kaggle-gemma]: https://www.kaggle.com/models/google/gemma-3\n[vertex-mg-gemma3]: https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3\n[terms]: https://ai.google.dev/gemma/terms\n[safety-policies]: https://ai.google/static/documents/ai-responsibility-update-published-february-2025.pdf\n[prohibited-use]: https://ai.google.dev/gemma/prohibited_use_policy\n[tpu]: https://cloud.google.com/tpu/docs/intro-to-tpu\n[sustainability]: https://sustainability.google/operating-sustainably/\n[jax]: https://github.com/jax-ml/jax\n[ml-pathways]: https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture/\n[sustainability]: https://sustainability.google/operating-sustainably/\n[gemini-2-paper]: https://arxiv.org/abs/2312.11805\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"86:T4a1,\u003c% let pyStream = request.stream.toString()[0].toUpperCase() + request.stream.toString().slice(1) %\u003e\nimport requests, base64\n\ninvoke_url = \"https://integrate.api.nvidia.com/v1/chat/completions\"\nstream = \u003c%- pyStream %\u003e\n\u003c% let content = \"What is in this image?\" %\u003e\nwith open(\"image.png\", \"rb\") as f:\n image_b64 = base64.b64encode(f.read()).decode()\n\nassert len(image_b64) \u003c 180_000, \\\n \"To upload larger images, use the assets API (see docs)\"\n \u003c% content = content.replaceAll(\"'\", \"\\\\'\") + \" \u003cimg src=\\\"data:image/png;base64,{image_b64}\\\" /\u003e\"%\u003e\n\nheaders = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"text/event-stream\" if stream else \"application/json\"\n}\n\npayload = {\n \"model\": '\u003c%- request.model %\u003e',\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": f'\u003c%- content %\u003e'\n }\n ],\n \"max_tokens\": \u003c%- request.max_tokens %\u003e,\n \"temperature\": \u003c%- request.temperature.toFixed(2) %\u003e,\n \"top_p\": \u003c%- request.top_p.toFixed(2) %\u003e,\n \"stream\": stream\n}\n\nresponse = requests.post(invoke_url, headers=headers, json=payload)\n\nif stream:\n for line in response.iter_lines():\n if line:\n print(line.decode(\"utf-8\"))\nelse:\n print(response.json())\n87:T550,import axios from 'axios';\nimport { readFile } from 'node:fs/promises';\n\nconst invokeUrl = \"https://integrate.api.nvidia.com/v1/chat/completions\";\nconst stream = \u003c%- request.stream %\u003e;\n\nconst headers = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": stream ? \"text/event-stream\" : \"application/json\"\n};\n\u003c% content = \"What is in this image? \u003cimg src=\\\"data:image/png;base64,${imageB64}\\\" /\u003e\" %\u003e\nreadFile(\"image.png\")\n .then(data =\u003e {\n const imageB64 = Buffer.from(data).toString('base64');\n if (imageB64.length \u003e 180_000) {\n throw new Error(\"To upload larger images, use the assets API (see docs)\");\n }\n\n const payload = {\n \"model\": `\u003c%- request.model %\u003e`,\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": `\u003c%- content %\u003e`\n }\n ],\n \"max_tokens\": \u003c%- request.max_tokens %"])</script><script>self.__next_f.push([1,"\u003e,\n \"temperature\": \u003c%- request.temperature.toFixed(2) %\u003e,\n \"top_p\": \u003c%- request.top_p.toFixed(2) %\u003e,\n \"stream\": stream\n };\n\n return axios.post(invokeUrl, payload, { headers: headers, responseType: stream ? 'stream' : 'json' });\n })\n .then(response =\u003e {\n if (stream) {\n response.data.on('data', (chunk) =\u003e {\n console.log(chunk.toString());\n });\n } else {\n console.log(JSON.stringify(response.data));\n }\n })\n .catch(error =\u003e {\n console.error(error);\n });\n88:T9328,"])</script><script>self.__next_f.push([1,"# Overview\n\n## Description: \nPhi-4-multimodal-instruct is a lightweight open multimodal foundation model that leverages the language, vision, and speech research and datasets used for Phi-3.5 and 4.0 models. The model processes text, image, and audio inputs, generating text outputs, and comes with 128K token context length. The model underwent an enhancement process, incorporating both supervised fine-tuning, and direct preference optimization to support precise instruction adherence and safety measures.\n\nThis model is ready for commercial/non-commercial use.\n\n## Third-Party Community Consideration\nThis model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA [Phi-4-Multimodal-Instruct](https://huggingface.co/microsoft/Phi-4-multimodal-instruct/blob/main/README.md).\n### License/Terms of Use:\nGOVERNING TERMS: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). Additional Information: [MIT License](https://huggingface.co/microsoft/Phi-4-multimodal-instruct/blob/main/LICENSE).\n\n### Deployment Geography:\nGlobal\n\n### Release Date:\nFebruary 2025\n\n## Reference(s):\n[Phi-4-Multimodal-Instruct Model Card](https://huggingface.co/microsoft/Phi-4-multimodal-instruct/blob/main/README.md)\n\n## Intended Use\n### Primary use cases:\nThe model is intended for broad multilingual and multimodal commercial and research use. It provides uses for general purpose AI systems and applications which require memory/compute constrained environments, latency bound scenarios, strong reasoning, function and tool calling, general image understanding, optical character recognition, and chart and table understanding. \n\nThe model is designed to accelerate research on language and multimodal models, for use as a building block for generative AI powered features.\n\n### Out-of-scope use cases\nThe model is not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models and multimodal models, as well as performance differences across languages, as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios. Developers should be aware of and adhere to applicable laws or regulations (including but not limited to privacy, trade compliance laws, etc.) that are relevant to their use case.\n\n**Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under.**\n\n## Release Notes\nThis release of Phi-4-multimodal-instruct is based on valuable user feedback from the Phi-3 series. Previously, users could use a speech recognition model to talk to the Mini and Vision models. To achieve this, users needed to use a pipeline of two models: one model to transcribe the audio to text, and another model for the language or vision tasks. This pipeline means that the core model was not provided the full breadth of input information – e.g. cannot directly observe multiple speakers, background noises, jointly align speech, vision, language information at the same time on the same representation space.\n\nWith Phi-4-multimodal-instruct, a single new open model has been trained across text, vision, and audio, meaning that all inputs and outputs are processed by the same neural network. The model employed new architecture, larger vocabulary for efficiency, multilingual, and multimodal support, and better post-training techniques were used for instruction following and function calling, as well as additional data leading to substantial gains on key multimodal capabilities.\n\nIt is anticipated that Phi-4-multimodal-instruct will greatly benefit app developers and various use cases. The enthusiastic support for the Phi-4 series is greatly appreciated. Feedback on Phi-4-multimodal-instruct is welcomed and crucial to the model’s evolution and improvement. Thank you for being part of this journey!\n\n## Model Architecture:\n**Architecture Type:** Phi-4-multimodal-instruct has 5.6B parameters and is a multimodal transformer model. The model has the pretrained Phi-4-mini as the backbone language model, and the advanced encoders and adapters of vision and speech\n\n## Input: \n**Input Type(s):** Text, Image, Audio \u003cbr\u003e\n**Input Format(s):** String, [.png, .jpg, .jpeg], [.mp3, .wav] \u003cbr\u003e\n**Input Parameters:** [1D, 2D] \u003cbr\u003e\n**Other Properties Related to Input:** Languages in training data | Arabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Hebrew, Hungarian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian \u003cbr\u003e \n**Note that these are for TEXT only. There is limited language support for IMAGE and AUDIO modalities.** \u003cbr\u003e \n* Vision: English \u003cbr\u003e \n* Audio: English, Chinese, German, French, Italian, Japanese, Spanish, Portuguese \u003cbr\u003e\n\n### Vision\n* Any common RGB/gray image format (e.g., (\".jpg\", \".jpeg\", \".png\", \".ppm\", \".bmp\", \".pgm\", \".tif\", \".tiff\", \".webp\")) can be supported.\n* Resolution depends on the GPU memory size. Higher resolution and more images will produce more tokens, thus using more GPU memory. During training, 64 crops can be supported. If it is a square image, the resolution would be around (8*448 by 8*448). For multiple-images, at most 64 frames can be supported, but with more frames as input, the resolution of each frame needs to be reduced to fit in the memory.\n### Audio\n\n* Any audio format that can be loaded by soundfile package should be supported.\n* To keep the satisfactory performance, maximum audio length is suggested to be 40 seconds. For summarization tasks, the maximum audio length is suggested to 30 minutes.\n\n\n## Output: \n**Output Type(s):** Text \u003cbr\u003e\n**Output Format(s):** String \u003cbr\u003e\n**Output Parameters:** 1D \u003cbr\u003e\n\n**Supported Hardware Microarchitecture Compatibility:**\n* NVIDIA Ampere \u003cbr\u003e\n* NVIDIA Hopper\n\n**[Preferred/Supported] Operating System(s):**\n* Linux\n\n## Model Version(s):\nPhi-4-multimodal-instruct v1.0\n\n## Training, Testing, and Evaluation Datasets:\n**Data Collection Methods**: [Hybrid: Automated, Human, Synthetic] \u003cbr\u003e\n**GPUS**: 512 A100-80G \u003cbr\u003e\n**Training Time**: 28 days \u003cbr\u003e\n**Training Data**: 5T text tokens, 2.3M speech hours, and 1.1T image-text token \u003cbr\u003e\n**Training Dates**: Trained between December 2024 and January 2025 \u003cbr\u003e\n**Status**: This is a static model trained on offline datasets with the cutoff date of June 2024 for publicly available data. \u003cbr\u003e\n**Languages in training data**: Arabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Hebrew, Hungarian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian \u003cbr\u003e **Note that these are for TEXT only. There is limited language support for IMAGE and AUDIO modalities. \u003cbr\u003e Vision: English\u003cbr\u003eAudio: English, Chinese, German, French, Italian, Japanese, Spanish, Portuguese**\n\n## Data Overview\nPhi-4-multimodal-instruct’s training data includes a wide variety of sources, totaling 5 trillion text tokens, and is a combination of:\n1. Publicly available documents filtered for quality, selected high-quality educational data, and code\n2. Newly created synthetic, “textbook-like” data for the purpose of teaching math, coding, common sense reasoning, general knowledge of the world (e.g., science, daily activities, theory of mind, etc.)\n3. High quality human labeled data in chat format\n4. Selected high-quality image-text interleave data\n5. Synthetic and publicly available image, multi-image, and video data\n6. Anonymized in-house speech-text pair data with strong/weak transcriptions\n7. Selected high-quality publicly available and anonymized in-house speech data with task-specific supervisions\n8. Selected synthetic speech data\n9. Synthetic vision-speech data\n\nFocus was placed on the quality of data that could potentially improve the reasoning ability for the model, and the publicly available documents were filtered to contain a preferred level of knowledge. As an example, the result of a game in premier league on a particular day might be good training data for large foundation models, but such information was removed for the Phi-4-multimodal-instruct to leave more model capacity for reasoning for the model’s small size. The data collection process involved sourcing information from publicly available documents, with a focus on filtering out undesirable documents and images. To safeguard privacy, image and text data sources were filtered to remove or scrub potentially personal data from the training data.\n\nThe decontamination process involved normalizing and tokenizing the dataset, then generating and comparing n-grams between the target dataset and benchmark datasets. Samples with matching n-grams above a threshold were flagged as contaminated and removed from the dataset. A detailed contamination report was generated, summarizing the matched text, matching ratio, and filtered results for further analysis.\n\n## Safety\n\n### Approach\nThe Phi-4 family of models has adopted a robust safety post-training approach. This approach leverages a variety of both open-source and in-house generated datasets. The overall technique employed for safety alignment is a combination of SFT (Supervised Fine-Tuning), DPO (Direct Preference Optimization), and RLHF (Reinforcement Learning from Human Feedback) approaches by utilizing human-labeled and synthetic English-language datasets, including publicly available datasets focusing on helpfulness and harmlessness, as well as various questions and answers targeted to multiple safety categories. For non-English languages, existing datasets were extended via machine translation. Speech Safety datasets were generated by running Text Safety datasets through Azure TTS (Text-To-Speech) Service, for both English and non-English languages. Vision (text \u0026 images) Safety datasets were created to cover harm categories identified both in public and internal multi-modal RAI datasets.\n\n### Safety Evaluation and Red-Teaming\nVarious evaluation techniques including red teaming, adversarial conversation simulations, and multilingual safety evaluation benchmark datasets were leveraged to evaluate Phi-4 models’ propensity to produce undesirable outputs across multiple languages and risk categories. Several approaches were used to compensate for the limitations of one approach alone. Findings across the various evaluation methods indicate that safety post-training that was done as detailed in the Phi 3 Safety Post-Training paper had a positive impact across multiple languages and risk categories as observed by refusal rates (refusal to output undesirable outputs) and robustness to jailbreak techniques. Details on prior red team evaluations across Phi models can be found in the Phi 3 Safety Post-Training paper. For this release, the red teaming effort focused on the newest Audio input modality and on the following safety areas: harmful content, self-injury risks, and exploits. The model was found to be more susceptible to providing undesirable outputs when attacked with context manipulation or persuasive techniques. These findings applied to all languages, with the persuasive techniques mostly affecting French and Italian. This highlights the need for industry-wide investment in the development of high-quality safety evaluation datasets across multiple languages, including low resource languages, and risk areas that account for cultural nuances where those languages are spoken.\n\n### Vision Safety Evaluation\nTo assess model safety in scenarios involving both text and images, Microsoft’s Azure AI Evaluation SDK was utilized. This tool facilitates the simulation of single-turn conversations with the target model by providing prompt text and images designed to incite harmful responses. The target model's responses are subsequently evaluated by a capable model across multiple harm categories, including violence, sexual content, self-harm, hateful and unfair content, with each response scored based on the severity of the harm identified. The evaluation results were compared with those of Phi-3.5-Vision and open-source models of comparable size. In addition, we ran both an internal and the public RTVLM and VLGuard multi-modal (text \u0026 vision) RAI benchmarks, once again comparing scores with Phi-3.5-Vision and open-source models of comparable size. However, the model may be susceptible to language-specific attack prompts and cultural context.\n\n### Audio Safety Evaluation\nIn addition to extensive red teaming, the Safety of the model was assessed through three distinct evaluations. First, as performed with Text and Vision inputs, Microsoft’s Azure AI Evaluation SDK was leveraged to detect the presence of harmful content in the model’s responses to Speech prompts. Second, Microsoft’s Speech Fairness evaluation was run to verify that Speech-To-Text transcription worked well across a variety of demographics. Third, we proposed and evaluated a mitigation approach via a system message to help prevent the model from inferring sensitive attributes (such as gender, sexual orientation, profession, medical condition, etc...) from the voice of a user.\n\n## Model Quality\n\nTo understand the capabilities, Phi-4-multimodal-instruct was compared with a set of models over a variety of benchmarks using an internal benchmark platform. Users can refer to the Phi-4-Mini model card for details of language benchmarks. Below is a high-level overview of the model quality on representative speech and vision benchmarks:\n\n### Speech Benchmarks\n\nPhi-4-multimodal-instruct demonstrated strong performance in speech tasks:\n- Surpassed expert ASR model WhisperV3 and ST models SeamlessM4T-v2-Large in automatic speech recognition (ASR) and speech translation (ST).\n- Ranked number 1 on the Huggingface OpenASR leaderboard with a word error rate of 6.14% compared to the current best model at 6.5% as of February 18, 2025.\n- First open-sourced model capable of performing speech summarization, with performance close to GPT4o.\n- Exhibited a gap with closed models like Gemini-2.0-Flash and GPT-4o-realtime-preview on the speech QA task. Efforts are ongoing to improve this capability in future iterations.\n\n### Vision Benchmarks\n\n#### Vision-Speech Tasks\n\nPhi-4-multimodal-instruct can process both image and audio together. The table below shows the model quality when the input query for vision content is synthetic speech on chart/table understanding and document reasoning tasks. Compared to other state-of-the-art omni models, Phi-4-multimodal-instruct achieves stronger performance on multiple benchmarks.\n\n| Benchmarks | Phi-4-multimodal-instruct | InternOmni-7B | Gemini-2.0-Flash-Lite-prv-02-05 | Gemini-2.0-Flash | Gemini-1.5-Pro |\n|------------|----------------------------|---------------|---------------------------------|------------------|----------------|\n| s_AI2D | 68.9 | 53.9 | 62.0 | 69.4 | 67.7 |\n| s_ChartQA | 69.0 | 56.1 | 35.5 | 51.3 | 46.9 |\n| s_DocVQA | 87.3 | 79.9 | 76.0 | 80.3 | 78.2 |\n| s_InfoVQA | 63.7 | 60.3 | 59.4 | 63.6 | 66.1 |\n| **Average**| **72.2** | **62.6** | **58.2** | **66.2** | **64.7** |\n\n#### Vision Tasks\n\nTo understand the vision capabilities, Phi-4-multimodal-instruct was compared with a set of models over a variety of zero-shot benchmarks using an internal benchmark platform. Below is a high-level overview of the model quality on representative benchmarks:\n\n| Dataset | Phi-4-multimodal-ins | Phi-3.5-vision-ins | Qwen 2.5-VL-3B-ins | Intern VL 2.5-4B | Qwen 2.5-VL-7B-ins | Intern VL 2.5-8B | Gemini 2.0-Flash Lite-prv-0205 | Gemini2.0-Flash | Claude-3.5-Sonnet-2024-10-22 | Gpt-4o-2024-11-20 |\n|------------------------------|----------------------|--------------------|--------------------|------------------|--------------------|------------------|---------------------------------|-----------------|--------------------------------|-------------------|\n| Popular aggregated benchmark | 55.1 | 43.0 | 47.0 | 48.3 | 51.8 | 50.6 | 54.1 | 64.7 | 55.8 | 61.7 |\n| MMBench (dev-en) | 86.7 | 81.9 | 84.3 | 86.8 | 87.8 | 88.2 | 85.0 | 90.0 | 86.7 | 89.0 |\n| MMMU-Pro (std / vision) | 38.5 | 21.8 | 29.9 | 32.4 | 38.7 | 34.4 | 45.1 | 54.4 | 54.3 | 53.0 |\n| ScienceQA Visual (img-test) | 97.5 | 91.3 | 79.4 | 96.2 | 87.7 | 97.3 | 85.0 | 88.3 | 81.2 | 88.2 |\n| MathVista (testmini) | 62.4 | 43.9 | 60.8 | 51.2 | 67.8 | 56.7 | 57.6 | 47.2 | 56.9 | 56.1 |\n| InterGPS | 48.6 | 36.3 | 48.3 | 53.7 | 52.7 | 54.1 | 57.9 | 65.4 | 47.1 | 49.1 |\n| AI2D | 82.3 | 78.1 | 78.4 | 80.0 | 82.6 | 83.0 | 77.6 | 82.1 | 70.6 | 83.8 |\n| ChartQA | 81.4 | 81.8 | 80.0 | 79.1 | 85.0 | 81.0 | 73.0 | 79.0 | 78.4 | 75.1 |\n| DocVQA | 93.2 | 69.3 | 93.9 | 91.6 | 95.7 | 93.0 | 91.2 | 92.1 | 95.2 | 90.9 |\n| InfoVQA | 72.7 | 36.6 | 77.1 | 72.1 | 82.6 | 77.6 | 73.0 | 77.8 | 74.3 | 71.9 |\n| TextVQA (val) | 75.6 | 72.0 | 76.8 | 70.9 | 77.7 | 74.8 | 72.9 | 74.4 | 58.6 | 73.1 |\n| OCR Bench | 84.4 | 63.8 | 82.2 | 71.6 | 87.7 | 74.8 | 75.7 | 81.0 | 77.0 | 77.7 |\n| POPE | 85.6 | 86.1 | 87.9 | 89.4 | 87.5 | 89.1 | 87.5 | 88.0 | 82.6 | 86.5 |\n| BLINK | 61.3 | 57.0 | 48.1 | 51.2 | 55.3 | 52.5 | 59.3 | 64.0 | 56.9 | 62.4 |\n| Video MME (16 frames) | 55.0 | 50.8 | 56.5 | 57.3 | 58.2 | 58.7 | 58.8 | 65.5 | 60.2 | 68.2 |\n| **Average** | **72.0** | **60.9** | **68.7** | **68.8** | **73.3** | **71.1** | **70.2** | **74.3** | **69.1** | **72.4** |\n\nBelow are the comparison results on existing multi-image tasks. On average, Phi-4-multimodal-instruct outperforms competitor models of the same size and is competitive with much bigger models on multi-frame capabilities. BLINK is an aggregated benchmark with 14 visual tasks that humans can solve very quickly but are still hard for current multimodal LLMs.\n\n| Dataset | Phi-4-multimodal-instruct | Qwen2.5-VL-3B-Instruct | InternVL 2.5-4B | Qwen2.5-VL-7B-Instruct | InternVL 2.5-8B | Gemini-2.0-Flash-Lite-prv-02-05 | Gemini-2.0-Flash | Claude-3.5-Sonnet-2024-10-22 | Gpt-4o-2024-11-20 |\n|------------------------|---------------------------|------------------------|-----------------|------------------------|-----------------|---------------------------------|------------------|--------------------------------|-------------------|\n| Art Style | 86.3 | 58.1 | 59.8 | 65.0 | 65.0 | 76.9 | 76.9 | 68.4 | 73.5 |\n| Counting | 60.0 | 67.5 | 60.0 | 66.7 | 71.7 | 45.8 | 69.2 | 60.8 | 65.0 |\n| Forensic Detection | 90.2 | 34.8 | 22.0 | 43.9 | 37.9 | 31.8 | 74.2 | 63.6 | 71.2 |\n| Functional Correspondence | 30.0 | 20.0 | 26.9 | 22.3 | 27.7 | 48.5 | 53.1 | 34.6 | 42.3 |\n| IQ Test | 22.7 | 25.3 | 28.7 | 28.7 | 28.7 | 28.0 | 30.7 | 20.7 | 25.3 |\n| Jigsaw | 68.7 | 52.0 | 71.3 | 69.3 | 53.3 | 62.7 | 69.3 | 61.3 | 68.7 |\n| Multi-View Reasoning | 76.7 | 44.4 | 44.4 | 54.1 | 45.1 | 55.6 | 41.4 | 54.9 | 54.1 |\n| Object Localization | 52.5 | 55.7 | 53.3 | 55.7 | 58.2 | 63.9 | 67.2 | 58.2 | 65.6 |\n| Relative Depth | 69.4 | 68.5 | 68.5 | 80.6 | 76.6 | 81.5 | 72.6 | 66.1 | 73.4 |\n| Relative Reflectance | 26.9 | 38.8 | 38.8 | 32.8 | 38.8 | 33.6 | 34.3 | 38.1 | 38.1 |\n| Semantic Correspondence | 52.5 | 32.4 | 33.8 | 28.8 | 24.5 | 56.1 | 55.4 | 43.9 | 47.5 |\n| Spatial Relation | 72.7 | 80.4 | 86.0 | 88.8 | 86.7 | 74.1 | 79.0 | 74.8 | 83.2 |\n| Visual Correspondence | 67.4 | 28.5 | 39.5 | 50.0 | 44.2 | 84.9 | 91.3 | 72.7 | 82.6 |\n| Visual Similarity | 86.7 | 67.4 | 88.1 | 87.4 | 85.2 | 87.4 | 80.7 | 79.3 | 83.0 |\n| **Overall** | **61.3** | **48.1** | **51.2** | **55.3** | **52.5** | **59.3** | **64.0** | **56.9** | **62.4** |\n\n# Usage\n\n### Input Format\n\nGiven the nature of the training data, Phi-4-Multimodal-Instruct model is best suited for prompts using the chat format as follows:\n\n1. **Text Chat Format**\n\n This format is used for general conversation and instructions:\n ```\n \u003c|system|\u003eYou are a helpful assistant.\u003c|end|\u003e\u003c|user|\u003eHow to explain Internet for a medieval knight?\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n1. **Tool-enabled Function Call Format for Text**\n\n This format is used when the user wants the model to provide function calls based on the given tools. The user should provide the available tools in the system prompt, wrapped by `\u003c|tool|\u003e` and `\u003c|/tool|\u003e` tokens. The tools should be specified in JSON format, using a JSON dump structure. For example:\n ```\n \u003c|system|\u003eYou are a helpful assistant with some tools.\u003c|tool|\u003e\n [{\"name\": \"get_weather_updates\", \"description\": \"Fetches weather updates for a given city using the RapidAPI Weather API.\", \"parameters\": {\"city\": {\"description\": \"The name of the city for which to retrieve weather information.\", \"type\": \"str\", \"default\": \"London\"}}}]\n \u003c|/tool|\u003e\u003c|end|\u003e\u003c|user|\u003eWhat is the weather like in Paris today?\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n1. **Vision-Language Format**\n\n This format is used for conversation with image:\n ``` \n \u003c|user|\u003e\u003c|image_1|\u003eDescribe the image in detail.\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n For multiple images, the user needs to insert multiple image placeholders in the prompt as below:\n ```\n \u003c|user|\u003e\u003c|image_1|\u003e\u003c|image_2|\u003e\u003c|image_3|\u003eSummarize the content of the images.\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n1. **Speech-Language Format**\n\n This format is used for various speech and audio tasks:\n ```\n \u003c|user|\u003e\u003c|audio_1|\u003e{task prompt}\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n The task prompt can vary for different task.\n * **Automatic Speech Recognition:**\n ```\n \u003c|user|\u003e\u003c|audio_1|\u003eTranscribe the audio clip into text.\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n * **Automatic Speech Translation:**\n ```\n \u003c|user|\u003e\u003c|audio_1|\u003eTranslate the audio to {lang}.\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n * **Automatic Speech Translation with Chain-of-thought:**\n ```\n \u003c|user|\u003e\u003c|audio_1|\u003eTranscribe the audio to text, and then translate the audio to {lang}. Use \u003csep\u003e as a separator between the original transcript and the translation.\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n * **Spoken-query Question Answering:**\n ```\n \u003c|user|\u003e\u003c|audio_1|\u003e\u003c|end|\u003e\u003c|assistant|\u003e\n ```\n1. **Vision-Speech Format**\n\nThis format is used for conversation with image and audio. The audio may contain query related to the image:\n```\n\u003c|user|\u003e\u003c|image_1|\u003e\u003c|audio_1|\u003e\u003c|end|\u003e\u003c|assistant|\u003e\n```\nFor multiple images, the user needs to insert multiple image placeholders in the prompt as below:\n```\n\u003c|user|\u003e\u003c|image_1|\u003e\u003c|image_2|\u003e\u003c|image_3|\u003e\u003c|audio_1|\u003e\u003c|end|\u003e\u003c|assistant|\u003e\n```\n\n### Vision\n\n* Any common RGB/gray image format (e.g., (\".jpg\", \".jpeg\", \".png\", \".ppm\", \".bmp\", \".pgm\", \".tif\", \".tiff\", \".webp\")) can be supported.\n* Resolution depends on the GPU memory size. Higher resolution and more images will produce more tokens, thus using more GPU memory. During training, 64 crops can be supported. If it is a square image, the resolution would be around (8*448 by 8*448). For multiple-images, at most 64 frames can be supported, but with more frames as input, the resolution of each frame needs to be reduced to fit in the memory.\n\n### Audio\n\n* Any audio format that can be loaded by soundfile package should be supported.\n* To keep the satisfactory performance, maximum audio length is suggested to be 40 seconds. For summarization tasks, the maximum audio length is suggested to 30 minutes.\n\n### Loading the Model Locally\n\nAfter obtaining the Phi-4-Multimodal-Instruct model checkpoints, users can use this sample code for inference.\n\n```python\nimport requests\nimport torch\nimport os\nfrom PIL import Image\nimport soundfile\nfrom transformers import AutoModelForCausalLM, AutoProcessor, GenerationConfig,pipeline,AutoTokenizer\n\nprocessor = AutoProcessor.from_pretrained(model_path, trust_remote_code=True)\n\nmodel = AutoModelForCausalLM.from_pretrained(\n \"microsoft/Phi-4-Multimodal-Instruct\", \n device_map=\"cuda\", \n torch_dtype=\"auto\", \n trust_remote_code=True, \n_attn_implementation='flash_attention_2',\n).cuda()\n\ngeneration_config = GenerationConfig.from_pretrained(model_path, 'generation_config.json')\n\nuser_prompt = '\u003c|user|\u003e'\nassistant_prompt = '\u003c|assistant|\u003e'\nprompt_suffix = '\u003c|end|\u003e'\n\nprompt = f'{user_prompt}\u003c|image_1|\u003eWhat is shown in this image?{prompt_suffix}{assistant_prompt}'\nurl = 'https://www.ilankelman.org/stopsigns/australia.jpg'\nprint(f'\u003e\u003e\u003e Prompt\\n{prompt}')\nimage = Image.open(requests.get(url, stream=True).raw)\ninputs = processor(text=prompt, images=image, return_tensors='pt').to('cuda:0')\ngenerate_ids = model.generate(\n **inputs,\n max_new_tokens=1000,\n generation_config=generation_config,\n)\ngenerate_ids = generate_ids[:, inputs['input_ids'].shape[1] :]\nresponse = processor.batch_decode(\n generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False\n)[0]\nprint(f'\u003e\u003e\u003e Response\\n{response}')\n\n\nspeech_prompt = \"Transcribe the audio to text, and then translate the audio to French. Use \u003csep\u003e as a separator between the original transcript and the translation.\"\nprompt = f'{user_prompt}\u003c|audio_1|\u003e{speech_prompt}{prompt_suffix}{assistant_prompt}'\n\nprint(f'\u003e\u003e\u003e Prompt\\n{prompt}')\naudio = soundfile.read('https://voiceage.com/wbsamples/in_mono/Trailer.wav')\ninputs = processor(text=prompt, audios=[audio], return_tensors='pt').to('cuda:0')\ngenerate_ids = model.generate(\n **inputs,\n max_new_tokens=1000,\n generation_config=generation_config,\n)\ngenerate_ids = generate_ids[:, inputs['input_ids'].shape[1] :]\nresponse = processor.batch_decode(\n generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False\n)[0]\nprint(f'\u003e\u003e\u003e Response\\n{response}')\n```\n\n## Inference:\n**Engine:** vLLM \u003cbr\u003e\n**Test Hardware:** NVIDIA H100\n\n## Responsible AI Considerations\n\nLike other language models, the Phi family of models can potentially behave in ways that are unfair, unreliable, or offensive. Some of the limiting behaviors to be aware of include: \n\n- **Quality of Service:** The Phi models are trained primarily on English language content across text, speech, and visual inputs, with some additional multilingual coverage. Performance may vary significantly across different modalities and languages:\n - **Text:** Languages other than English will experience reduced performance, with varying levels of degradation across different non-English languages. English language varieties with less representation in the training data may perform worse than standard American English.\n - **Speech:** Speech recognition and processing shows similar language-based performance patterns, with optimal performance for standard American English accents and pronunciations. Other English accents, dialects, and non-English languages may experience lower recognition accuracy and response quality. Background noise, audio quality, and speaking speed can further impact performance.\n - **Vision:** Visual processing capabilities may be influenced by cultural and geographical biases in the training data. The model may show reduced performance when analyzing images containing text in non-English languages or visual elements more commonly found in non-Western contexts. Image quality, lighting conditions, and composition can also affect processing accuracy.\n- **Multilingual performance and safety gaps:** We believe it is important to make language models more widely available across different languages, but the Phi 4 models still exhibit challenges common across multilingual releases. As with any deployment of LLMs, developers will be better positioned to test for performance or safety gaps for their linguistic and cultural context and customize the model with additional fine-tuning and appropriate safeguards.\n- **Representation of Harms \u0026 Perpetuation of Stereotypes:** These models can over- or under-represent groups of people, erase representation of some groups, or reinforce demeaning or negative stereotypes. Despite safety post-training, these limitations may still be present due to differing levels of representation of different groups, cultural contexts, or prevalence of examples of negative stereotypes in training data that reflect real-world patterns and societal biases.\n- **Inappropriate or Offensive Content:** These models may produce other types of inappropriate or offensive content, which may make it inappropriate to deploy for sensitive contexts without additional mitigations that are specific to the case.\n- **Information Reliability:** Language models can generate nonsensical content or fabricate content that might sound reasonable but is inaccurate or outdated.\n- **Limited Scope for Code:** The majority of Phi 4 training data is based in Python and uses common packages such as \"typing, math, random, collections, datetime, itertools\". If the model generates Python scripts that utilize other packages or scripts in other languages, it is strongly recommended that users manually verify all API uses.\n- **Long Conversation:** Phi 4 models, like other models, can in some cases generate responses that are repetitive, unhelpful, or inconsistent in very long chat sessions in both English and non-English languages. Developers are encouraged to place appropriate mitigations, like limiting conversation turns to account for the possible conversational drift.\n\nDevelopers should apply responsible AI best practices, including mapping, measuring, and mitigating risks associated with their specific use case and cultural, linguistic context. Phi 4 family of models are general purpose models. As developers plan to deploy these models for specific use cases, they are encouraged to fine-tune the models for their use case and leverage the models as part of broader AI systems with language-specific safeguards in place. Important areas for consideration include:\n\n- **Allocation:** Models may not be suitable for scenarios that could have consequential impact on legal status or the allocation of resources or life opportunities (ex: housing, employment, credit, etc.) without further assessments and additional debiasing techniques.\n- **High-Risk Scenarios:** Developers should assess the suitability of using models in high-risk scenarios where unfair, unreliable or offensive outputs might be extremely costly or lead to harm. This includes providing advice in sensitive or expert domains where accuracy and reliability are critical (ex: legal or health advice). Additional safeguards should be implemented at the application level according to the deployment context.\n- **Misinformation:** Models may produce inaccurate information. Developers should follow transparency best practices and inform end-users they are interacting with an AI system. At the application level, developers can build feedback mechanisms and pipelines to ground responses in use-case specific, contextual information, a technique known as Retrieval Augmented Generation (RAG).\n- **Generation of Harmful Content:** Developers should assess outputs for their context and use available safety classifiers or custom solutions appropriate for their use case.\n- **Misuse:** Other forms of misuse such as fraud, spam, or malware production may be possible, and developers should ensure that their applications do not violate applicable laws and regulations.\n\n## Ethical Considerations:\nEthical considerations and guidelines. NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"89:T557,\u003c% let pyStream = request.stream.toString()[0].toUpperCase() + request.stream.toString().slice(1) %\u003e\nimport requests, base64\n\ninvoke_url = \"https://integrate.api.nvidia.com/v1/chat/completions\"\nstream = \u003c%- pyStream %\u003e\n\u003c% let content = \"Answer the spoken query about the image.\" %\u003e\nwith open(\"image.png\", \"rb\") as f:\n image_b64 = base64.b64encode(f.read()).decode()\nwith open(\"audio.wav\", \"rb\") as f:\n audio_b64 = base64.b64encode(f.read()).decode()\n\nassert len(image_b64) + len(audio_b64) \u003c 180_000, \\\n \"To upload larger images and/or audios, use the assets API (see docs)\"\n\u003c% content = content.replaceAll(\"'\", \"\\\\'\") + \"\u003cimg src=\\\"data:image/png;base64,{image_b64}\\\" /\u003e\u003caudio src=\\\"data:audio/wav;base64,{audio_b64}\\\" /\u003e\"%\u003e\n\nheaders = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"text/event-stream\" if stream else \"application/json\"\n}\n\npayload = {\n \"model\": '\u003c%- request.model %\u003e',\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": f'\u003c%- content %\u003e'\n }\n ],\n \"max_tokens\": \u003c%- request.max_tokens %\u003e,\n \"temperature\": \u003c%- request.temperature.toFixed(2) %\u003e,\n \"top_p\": \u003c%- request.top_p.toFixed(2) %\u003e,\n \"stream\": stream\n}\n\nresponse = requests.post(invoke_url, headers=headers, json=payload)\n\nif stream:\n for line in response.iter_lines():\n if line:\n print(line.decode(\"utf-8\"))\nelse:\n print(response.json())\n8a:T63e,import axios from 'axios';\nimport { readFile } from 'node:fs/promises';\n\nconst invokeUrl = \"https://integrate.api.nvidia.com/v1/chat/completions\";\nconst stream = \u003c%- request.stream %\u003e;\n\nconst headers = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": stream ? \"text/event-stream\" : \"application/json\"\n};\n\u003c% content = \"Answer the spoken query about the image.\u003cimg src=\\\"data:image/png;base64,${imageB64}\\\" /\u003e\u003caudio src=\\\"data:audio/wav;base64,${audioB64}\\\" /\u003e\" %\u003e\nPromise.all([\n readFile(\"image.png\"),\n readFile(\"audio.wav\")\n])\n .then(([imageData, audioData]) =\u003e {\n const imageB64 = Buffer.from(imageData).toString('base64');\n const audioB64 = Buf"])</script><script>self.__next_f.push([1,"fer.from(audioData).toString('base64');\n \n if (imageB64.length + audioB64.length \u003e 180_000) {\n throw new Error(\"To upload larger images and/or audios, use the assets API (see docs)\");\n }\n\n const payload = {\n \"model\": `\u003c%- request.model %\u003e`,\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": `\u003c%- content %\u003e`\n }\n ],\n \"max_tokens\": \u003c%- request.max_tokens %\u003e,\n \"temperature\": \u003c%- request.temperature.toFixed(2) %\u003e,\n \"top_p\": \u003c%- request.top_p.toFixed(2) %\u003e,\n \"stream\": stream\n };\n\n return axios.post(invokeUrl, payload, { headers: headers, responseType: stream ? 'stream' : 'json' });\n })\n .then(response =\u003e {\n if (stream) {\n response.data.on('data', (chunk) =\u003e {\n console.log(chunk.toString());\n });\n } else {\n console.log(JSON.stringify(response.data));\n }\n })\n .catch(error =\u003e {\n console.error(error);\n });\n8b:T24d3,"])</script><script>self.__next_f.push([1,"# **Cosmos-Predict1**: A Suite of Diffusion-based World Foundation Models\n\n[**Cosmos**](https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6) | [**Code**](https://github.com/NVIDIA/Cosmos) | [**Paper**](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai)\n\n\n# Model Overview\n\n## Description:\n**Cosmos World Foundation Models**: A family of highly performant pre-trained world foundation models purpose-built for generating physics-aware videos and world states for physical AI development.\n\nThe Cosmos diffusion models are a collection of diffusion based world foundation models that generate dynamic, high quality videos from text, image, or video inputs. It can serve as the building block for various applications or research that are related to world generation. The models are ready for commercial use under NVIDIA Open Model license agreement.\n\n**Model Developer**: NVIDIA\n\n## Model Versions\n\nIn Cosmos 1.0 release, the Cosmos Diffusion WFM family includes the following models:\n- [Cosmos-Predict1-7B-Text2World](https://huggingface.co/nvidia/Cosmos-Predict1-7B-Text2World)\n - Given a text description, predict an output video of 121 frames.\n- [Cosmos-Predict1-14B-Text2World](https://huggingface.co/nvidia/Cosmos-Predict1-14B-Text2World)\n - Given a text description, predict an output video of 121 frames.\n- [Cosmos-Predict1-7B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-7B-Video2World)\n - Given a text description and an image as the first frame, predict the future 120 frames.\n- [Cosmos-Predict1-14B-Video2World](https://huggingface.co/nvidia/Cosmos-Predict1-14B-Video2World)\n - Given a text description and an image as the first frame, predict the future 120 frames.\n\n\n### License:\nThis model is released under the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). For a custom license, please contact [cosmos-license@nvidia.com](mailto:cosmos-license@nvidia.com).\n\nUnder the NVIDIA Open Model License, NVIDIA confirms:\n\n* Models are commercially usable.\n* You are free to create and distribute Derivative Models.\n* NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.\n\n**Important Note**: If you bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or\nassociated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained\nin the Model, your rights under [NVIDIA Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/) will automatically terminate.\n\n## Model Architecture:\n\n**Cosmos-Predict1-7B-Text2World** and **Cosmos-Predict1-7B-Video2World** are diffusion transformer models designed for video denoising in the latent space. The network is composed of interleaved self-attention, cross-attention and feedforward layers as its building blocks. The cross-attention layers allow the model to condition on input text throughout the denoising process. Before each layers, adaptive layer normalization is applied to embed the time information for denoising. When image or video is provided as input, their latent frames are concatenated with the generated frames along the temporal dimension. Augment noise is added to conditional latent frames to bridge the training and inference gap.\n\n## Cosmos-Predict1-7B-Text2World Input/Output Specifications\n\n* **Input**\n\n * **Input Type(s)**: Text\n * **Input Format(s)**: String\n * **Input Parameters**: One-dimensional (1D)\n * **Other Properties Related to Input**:\n * The input string should contain fewer than 300 words and should provide descriptive content for world generation, such as a scene description, key objects or characters, background, and any specific actions or motions to be depicted within the 5-second duration.\n\n* **Output**\n * **Output Type(s)**: Video\n * **Output Format(s)**: mp4\n * **Output Parameters**: Three-dimensional (3D)\n * **Other Properties Related to Output**: The generated video will be a 5-second clip with a resolution of 1280x704 pixels at 24 frames per second (fps). The content of the video will visualize the input text description as a short animated scene, capturing the main elements mentioned in the input within the time constraints.\n\n## Cosmos-Predict1-7B-Video2World Input/Output Specifications\n\n* **Input**\n\n * **Input Type(s)**: Text+Image, Text+Video\n * **Input Format(s)**:\n * Text: String\n * Image: jpg, png, jpeg, webp\n * Video: mp4\n * **Input Parameters**:\n * Text: One-dimensional (1D)\n * Image: Two-dimensional (2D)\n * Video: Three-dimensional (3D)\n * **Other Properties Related to Input**:\n * The input string should contain fewer than 300 words and should provide descriptive content for world generation, such as a scene description, key objects or characters, background, and any specific actions or motions to be depicted within the 5-second duration.\n * The input image should be of 1280x704 resolution.\n * The input video should be of 1280x704 resolution and 9 input frames.\n\n* **Output**\n * **Output Type(s)**: Video\n * **Output Format(s)**: mp4\n * **Output Parameters**: Three-dimensional (3D)\n * **Other Properties Related to Output**: The generated video will be a 5-second clip with a resolution of 1280x704 pixels at 24 frames per second (fps). The content of the video will use the provided image as the first frame and visualize the input text description as a short animated scene, capturing the main elements mentioned in the input within the time constraints.\n\n## Software Integration\n**Runtime Engine(s):**\n* [Cosmos](https://github.com/NVIDIA/Cosmos)\n\n**Supported Hardware Microarchitecture Compatibility:**\n* NVIDIA Blackwell\n* NVIDIA Hopper\n* NVIDIA Ampere\n\n**Note**: We have only tested inference with BF16 precision.\n\n\n\n**Operating System(s):**\n* Linux (We have not tested on other operating systems.)\n\n\n# Usage\n\n* See [Cosmos](https://github.com/NVIDIA/Cosmos) for details.\n\n\n# Evaluation\n\nPlease see our [technical paper](https://research.nvidia.com/publication/2025-01_cosmos-world-foundation-model-platform-physical-ai) for detailed evaluations.\n\n## Inference Time and GPU Memory Usage\n\nThe numbers provided below may vary depending on system specs and are for reference only.\n\nWe report the maximum observed GPU memory usage during end-to-end inference. Additionally, we offer a series of model offloading strategies to help users manage GPU memory usage effectively.\n\nFor GPUs with limited memory (e.g., RTX 3090/4090 with 24 GB memory), we recommend fully offloading all models. For higher-end GPUs, users can select the most suitable offloading strategy considering the numbers provided below.\n\n### Cosmos-Predict1-7B-Text2World\n\n| Offloading Strategy | 7B Text2World | 14B Text2World |\n|-------------|---------|---------|\n| Offload prompt upsampler | 74.0 GB | \u003e 80.0 GB |\n| Offload prompt upsampler \u0026 guardrails | 57.1 GB | 70.5 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder | 38.5 GB | 51.9 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer | 38.3 GB | 51.7 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer \u0026 diffusion model | 24.4 GB | 39.0 GB |\n\nThe table below presents the end-to-end inference runtime on a single H100 GPU, excluding model initialization time.\n\n| 7B Text2World (offload prompt upsampler) | 14B Text2World (offload prompt upsampler, guardrails) |\n|---------|---------|\n| ~380 seconds | ~590 seconds |\n\n### Cosmos-Predict1-7B-Video2World\n\n| Offloading Strategy | 7B Video2World | 14B Video2World |\n|----------------------------------------------------------------------------------|---------|---------|\n| Offload prompt upsampler | 76.5 GB | \u003e 80.0 GB |\n| Offload prompt upsampler \u0026 guardrails | 59.9 GB | 73.3 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder | 41.3 GB | 54.8 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer | 41.1 GB | 54.5 GB |\n| Offload prompt upsampler \u0026 guardrails \u0026 T5 encoder \u0026 tokenizer \u0026 diffusion model | 27.3 GB | 39.0 GB |\n\nThe following table shows the end-to-end inference runtime on a single H100 GPU, excluding model initialization time:\n\n| 7B Video2World (offload prompt upsampler) | 14B Video2World (offload prompt upsampler, guardrails) |\n|---------|---------|\n| ~383 seconds | ~593 seconds |\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nFor more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety \u0026 Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"8c:T6cb,Field | Response\n:------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\nIntended Application \u0026 Domain: | World Generation\nModel Type: | Transformer\nIntended Users: | Physical AI developers\nOutput: | Videos\nDescribe how the model works: | Generates videos based on video inputs\nTechnical Limitations: | The model may not follow the video input accurately.\nVerified to have met prescribed NVIDIA quality standards: | Yes\nPerformance Metrics: | Quantitative and Qualitative Evaluation\nPotential Known Risks: | The model's output can generate all forms of videos, including what may be considered toxic, offensive, or indecent.\nLicensing: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)8d:T7ad,Field | Response\n:----------------------------------------------------------------------------------------------------------------------------------|:-----------------"])</script><script>self.__next_f.push([1,"------------------------------\nGeneratable or reverse engineerable personal information? | None Known\nProtected class data used to create this model? | None Known\nWas consent obtained for any personal data used? | None Known\nHow often is dataset reviewed? | Before Release\nIs a mechanism in place to honor data subject right of access or deletion of personal data? | Not Applicable\nIf personal data was collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable\nIf personal data was collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable\nIf personal data was collected for the development of this AI model, was it minimized to only what was required? | Not Applicable\nIs there provenance for all datasets used in training? | Yes\nDoes data labeling (annotation, metadata) comply with privacy laws? | Yes\nIs data compliant with data subject requests for data correction or removal, if such a request was made? | Not Applicable8e:T719,invoke_url='https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b'\nfetch_url_format='https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/'\n\nauthorization_header='Authorization: Bearer $NVIDIA_API_KEY'\naccept_header='Accept: application/json'\ncontent_type_header='Content-Type: application/json'\n\ndata='{\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\":"])</script><script>self.__next_f.push([1," [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on the industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n}'\n\nresponse=$(curl --silent -i -w \"\\n%{http_code}\" --request POST \\\n --url \"$invoke_url\" \\\n --header \"$authorization_header\" \\\n --header \"$accept_header\" \\\n --header \"$content_type_header\" \\\n --data \"$data\"\n)\n\nhttp_code=$(echo \"$response\" | tail -n 1)\nreq_id=$(echo \"$response\" | grep -i '^nvcf-reqid:' | awk '{print $2}' | tr -d '\\r')\n\nwhile [ \"$http_code\" -eq 202 ]; do\n response=$(curl --silent -i -w \"\\n%{http_code}\" --request GET \\\n --url \"$fetch_url_format$req_id\" \\\n --header \"$authorization_header\" \\\n --header \"$accept_header\" \\\n --header \"$content_type_header\" \\\n )\n\n http_code=$(echo \"$response\" | tail -n 1)\n req_id=$(echo \"$response\" | grep -i '^nvcf-reqid:' | awk '{print $2}' | tr -d '\\r')\ndone\n\nif [ \"$http_code\" -ne 302 ]; then\n echo \"invocation failed with status $http_code\" \u003e\u00262\n echo \"$response\" \u003e\u00262\n exit 1\nfi\n\ndownload_url=$(echo \"$response\" | grep -i '^location:' | awk '{print $2}' | tr -d '\\r')\ncurl -L --output result.zip \"$download_url\"\n8f:T5b2,import fs from \"fs\";\n\nconst invokeUrl = \"https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b\";\nconst fetchUrlFormat = \"https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/\";\n\nconst headers = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"application/json\",\n};\n\nconst payload = {\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\": [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on th"])</script><script>self.__next_f.push([1,"e industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n};\n\nlet response = await fetch(invokeUrl, {\n method: \"post\",\n body: JSON.stringify(payload),\n headers: { \"Content-Type\": \"application/json\", ...headers }\n});\n\nwhile (response.status === 202) {\n const requestId = response.headers.get(\"NVCF-REQID\");\n const fetchUrl = fetchUrlFormat + requestId;\n response = await fetch(fetchUrl, {\n method: \"get\",\n headers: headers\n });\n}\n\nif (response.status !== 200) {\n const errBody = await (await response.blob()).text();\n throw \"invocation failed with status \" + response.status + \" \" + errBody;\n}\n\nfs.writeFileSync('result.zip', Buffer.from(await response.arrayBuffer()));\n90:T4f1,import requests\n\ninvoke_url = \"https://ai.api.nvidia.com/v1/cosmos/nvidia/cosmos-predict1-7b\"\nfetch_url_format = \"https://api.nvcf.nvidia.com/v2/nvcf/pexec/status/\"\n\nheaders = {\n \"Authorization\": \"Bearer $NVIDIA_API_KEY\",\n \"Accept\": \"application/json\",\n}\n\npayload = {\n \"inputs\": [\n {\n \"name\": \"command\",\n \"shape\": [1],\n \"datatype\": \"BYTES\",\n \"data\": [\n \"text2world --prompt=\\\"A first person view from the perspective from a human sized robot as it works in a chemical plant. The robot has many boxes and supplies nearby on the industrial shelves. The camera on moving forward, at a height of 1m above the floor. Photorealistic views\\\"\"\n ]\n }\n ],\n \"outputs\": [\n {\n \"name\": \"status\",\n \"datatype\": \"BYTES\",\n \"shape\": [1]\n }\n ]\n}\n\n# re-use connections\nsession = requests.Session()\n\nresponse = session.post(invoke_url, headers=headers, json=payload)\n\nwhile response.status_code == 202:\n request_id = response.headers.get(\"NVCF-REQID\")\n fetch_url = fetch_url_format + request_id\n response = session.get(fetch_url, headers=headers)\nresponse = requests.post(invoke_url, headers=headers, json=payload)\n\nr"])</script><script>self.__next_f.push([1,"esponse.raise_for_status()\n\nwith open('result.zip', 'wb') as f:\n f.write(response.content)\n91:T3958,"])</script><script>self.__next_f.push([1,"# NVIDIA Cosmos\n\nCosmos is NVIDIA’s World Foundation Model Development Platform that provides the tools to either finetune existing models or train new models from scratch.\n\n## Cosmos Model Family\n\nCosmos World Foundation Models (WFM) are a family of highly-performant pre-trained models purpose-built for generating physics-aware videos used for training robots. With Cosmos, developers can simulate a world in which robots function and train them to act and react responsibly in the real world before actual deployment.\n\nCosmos WFMs currently contain four main types of models: NeMo Curator, Cosmos Tokenizer, Cosmos Guardrail, and Cosmos World Foundation Model. NeMo Curator is a video curation pipeline that takes raw video frames, splits them into meaningful segments, and annotates them with semantic tags, object labels, and scene descriptions. The annotated images are then fed into the Cosmos Tokenizer, which produces a sequence of tokens. This step reduces data dimensionality enabling Cosmos World Foundation Model to effectively handle large or complex inputs for training. Cosmos WFM then consumes the curated/annotated video segments and learns the underlying physics and visual dynamics from real world data. When queried, Cosmos WFM outputs new token sequences that are then decoded back into high-resolution and physically realistic synthetic videos. Cosmos WFMs are pretrained on large-scale video datasets to expose them to a broad range of visual experiences, enabling them to serve as generalists. To construct a specialized WFM developers are expected to fine-tune Cosmos WFM using additional data collected from a specific use case. This additional data will help adapt Cosmos WFM to this intended use case, ensuring it can perform optimally under real-world conditions.\n\n## Specific Risk Areas and Mitigations\n\nWFMs can produce unrealistic outputs, generate unsafe content or may inadvertently amplify societal biases reflected in their training data. Collectively, these risks underscore the need for technical measures to mitigate risk and careful evaluation before leveraging Cosmos WFM in real-world applications.\n\n### **Cosmos Guardrail**\n\n\nFor the safe use of our world foundation models, we develop a comprehensive guardrail system. Cosmos Guardrail consists of two stages: the pre-Guard and the post-Guard stage. The pre-Guard stage leverages [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0), which is a fine-tuned version of [Llama-Guard](https://ai.meta.com/research/publications/llama-guard-llm-based-input-output-safeguard-for-human-ai-conversations/) trained on [NVIDIA’s Aegis Content Safety Dataset](https://huggingface.co/datasets/nvidia/Aegis-AI-Content-Safety-Dataset-1.0) and a blocklist filter that performs a lemmatized and whole-word keyword search to block harmful prompts. It then further sanitizes the user prompt by processing it through the Cosmos Text2World Prompt Upsampler. The post-Guard stage blocks harmful visual outputs using a video content safety classifier and a face blur filter.\n![](https://assets.ngc.nvidia.com/products/api-catalog/cosmos/system-card-diagram.jpg)\n\n\nCosmos pre-Guard first uses a simple blocklist-based checker for unsafe keyword detection. This is designed to block explicitly harmful generations by doing a keyword search on the prompt against a hard-coded blocklist of a large corpus of explicit and objectionable words. Input words are lemmatized using [WordNetLemmatizer](https://www.nltk.org/api/nltk.stem.WordNetLemmatizer.html?highlight=wordnet), a tool that uses a lexical database of the English language to extract the root word from its variants. These lemmatized words are then compared to the words in the hard-coded blocklist, and the entire prompt is rejected if any profanity is found.\n\nAs the second line of defense, Cosmos pre-Guard uses [Aegis-AI-Content-Safety-LlamaGuard-LLM-Defensive-1.0](https://huggingface.co/nvidia/Aegis-AI-Content-Safety-LlamaGuard-Defensive-1.0) to detect unsafe content in semantically-complex prompts. Aegis is able to classify prompts into13 critical safety risk categories: violence, sexual, criminal planning, weapons, substance abuse, suicide, child sexual abuse material, hatred, harassment, threat, and profanity. If the input prompt is categorized as unsafe by this prompt filter, the video is not generated, and an error message is displayed. Any prompt that does not fall into the above categories is considered safe from the prompt-filtering standpoint.\n\nPrior to passing the prompt to the world generation models, the prompt is further augmented and indirectly sanitized via the Cosmos Text2World Prompt Upsampler. This is a bespoke model that not only compensates for the lack of specificity in the prompt but also steers clear of objectionable denotations or connotations.\n\nCosmos post-Guard is a vision-domain guardrail that is activated after the world content has been generated and comprises a video content safety filter and a face blur filter. Our video content safety filter used in the post-Guard stage has been trained on carefully-curated datasets and evaluated on human\\- annotated datasets created by Cosmos Red Team. To calibrate model outputs for the intended use case in the robotics and autonomous vehicle domains, we also automatically detect and blur all faces. We use RetinaFace, a state-of-the-art face detection model, to identify facial regions with high confidence scores. For any generated face region larger than 20 × 20 pixels, we apply pixelation to obscure features while preserving the overall scene composition. Note that by blurring all generated human faces in the video, potential biases based on age, gender, race and ethnicity in the output video are reduced.\n\n### **Balanced Datasets**\n\n\nCosmos WFM is trained using both proprietary and publicly available video datasets. We curated about 100M clips of videos ranging from 2 to 60 seconds from a 20M hour-long video collection. For each clip, we use a VLM ([13B-parameter VILA model](https://huggingface.co/Efficient-Large-Model/VILA-13b)) to provide a video caption per 256 frames. As our goal is to create a VLM that is able to generate physically realistic videos, we use the video captions to curate the training dataset to cover various physical applications:\n\n* Driving (11%),\n* Hand motion and object manipulation (16%),\n* Human motion and activity (10%),\n* Spatial awareness and navigation (16%),\n* First person point-of-view (8%),\n* Nature dynamics (20%),\n* Dynamic camera movements (8%),\n* Synthetically rendered (4%)\n* Others (7%)\n\nTo ensure effective distribution of the dataset we employ a taxonomy-based classifier to label video types and prune those that introduce unrealistic behaviors, such as purely animated or abstract patterns. Certain categories relevant to world foundation models (like human actions and interactions) are upsampled, while less critical ones (such as landscapes) are downsampled.\n\nA significant amount of the initial video data is either semantically redundant or contains different visual effects, which may induce unwanted artifacts in the generated videos if not appropriately handled. We therefore designed a sequence of data processing steps to find the most valuable parts of the raw videos for training. Shot boundary detection identifies where one shot ends and another begins, after which all footage is re-encoded into a uniform, high-quality MP4 format to ensure consistent loading and reduce codec discrepancies. The resulting video segments undergo several filtering processes. Motion filtering removes clips that are static or excessively shaky, and tags the remaining clips with camera motion types to enhance training signals. Visual quality filtering uses a video assessment model trained on [DOVER](https://github.com/VQAssessment/DOVER) to discard the bottom 15% in perceptual quality and applies an image aesthetic model exclude footage that is aesthetically poor. A deduplication step uses [InternVideo2](https://arxiv.org/abs/2403.15377) embeddings to identify near-duplicate content and preserves the highest-resolution version for minimal quality loss.\n\n## **Evaluation Methods**\n\nWe employ a dedicated red team to actively probe the system using both standard and adversarial examples that are collected in an internal attack prompt dataset. These video outputs are annotated by a team of expert annotators, who were specially trained for our task, to classify the generated video on a scale of 1-5 on multiple categories of harm related to the safety taxonomy. These annotations also specify the start and end-frames where the unsafe content is detected, thereby generating high-quality annotations. The red team also probed each guardrail component independently with targeted examples to identify weaknesses and improve performance in edge cases. As of the date of publication, the red team has tested and annotated over10, 000 distinct prompt-video pairs that were carefully crafted to cover a broad range of unsafe content. We separate out our safety testing into 4 categories:\n\n**Targeted unsafe testing**\n\nTargeted unsafe testing involves generating a corpus of manually curated unsafe prompts. These are intended to emulate common unsafe interactions that are performed by non-technical users of the system with basic or limited knowledge of multimodal AI attack vectors. These have a high likelihood of being caught by prompt filters, e.g. “Video of a naked person”.\n\n**Adversarial Attack Testing**\n\nAdversarial attack testing involves generating a corpus of unsafe prompts following the styles of AI attack published in literature. This type of testing will also leverage some (not all) of the prompts from content safety datasets like Aegis and automation tooling like [Garak](https://github.com/NVIDIA/garak).\n\n**Prompt Upsampler Toxicity**\n\nPrompt Upsampler Toxicity refers to a phenomenon where automated methods are used to “upsample” or expand a prompt by adding detail or context and inadvertently introduce unsafe content. Monitoring and mitigating Prompt Upsampler Toxicity ensures that content moderation systems remain effective throughout the entire generation pipeline, preserving user trust and upholding ethical standards.\n\n**Accidental Mishap Testing**\n\nAccidental mishap testing involves emulating the experience of a user prompting the model with a benign prompt, and getting unsafe content in return. This is the hardest category to test, since it does not have a fixed method or protocol for generation.\n\n## Governing Terms/Terms of Use\n\nAll Cosmos WFMs are deployed globally and are covered under NVIDIA’s [Open Model License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). This license agreement confirms that:\n\n* Models are commercially usable.\n* You are free to create and distribute derivative models.\n* NVIDIA does not claim ownership of any outputs generated using the models or derivative models.\n\nIf users bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperparameter, encryption, security, digital rights management, or authentication mechanism contained in the Model, the user’s rights under the NVIDIA Open Model License Agreement will automatically terminate. If users are interested in a custom license, they may contact cosmos-license@nvidia.com.\n\n## Deployment\n\nCosmos WFM is released under an open, permissive NVIDIA license, allowing users to download the model weights and run it on their own hardware. This means that developers can integrate the WFM into their existing workflows without dependency on external APIs. They can also tailor the model to specific domain needs, retrain or fine-tune the model with their private data. This approach fosters innovation especially for under-resourced stakeholders that cannot rely on paid services.\n\nOnce downloaded, NVIDIA has less visibility into how or where Cosmos is deployed, reducing opportunities to enforce content policies or guardrails. Downloadable models grant complete control to users, but also transfer responsibility to the users for preventing misuse, and implementing safety mechanisms, such as watermarking and content moderation. Watermarking in the context of WFMs is crucial to ensure traceability, and user awareness that generated content might not be authentic. Watermarks allow viewers and downstream users to identify AI-generated or AI-manipulated videos, helping prevent misinformation and misuse. Even though watermarking is typically the responsibility of the user, we still encourage the use of open-source libraries for watermarking by downstream users of Cosmos WFM. NVIDIA has actively promoted watermarking and has worked in consortiums and standards bodies to define common protocols for watermarking synthetic media.\n\nCosmos WFM is also hosted by NVIDIA at the NVIDIA API Catalog (build.nvidia.com) and accessible via a web-based user interface. In this case, NVIDIA manages infrastructure, updates, and safety features. End users with minimal machine learning expertise can harness powerful WFMs without worrying about infrastructure or setup. Hosted models give NVIDIA more oversight and moderation capabilities, for example:\n\n* Know Your Customer (KYC) and account verification ensures that users are who they claim to be, discouraging malicious actors and fostering accountability.\n* Usage monitoring securely records user activity and flags suspicious patterns, enabling traceability and compliance checks while helping identify harmful behavior.\n* Rate limiting prevents spamming and large-scale misuse, balancing computational resources and protecting against abuse or overwhelming the system.\n* Human review protocols provide an escalation path for questionable outputs or flagged user accounts. This is a dedicated moderation team for final decisions on content removals, user bans, or investigations.\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal supporting team to ensure this system meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\n## Getting Help/Support\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n"])</script><script>self.__next_f.push([1,"92:{\"name\":\"llama-3_3-nemotron-super-49b-v1\",\"type\":\"model\"}\n93:{\"name\":\"deepseek-r1\",\"type\":\"model\"}\n94:{\"name\":\"llama-3_1-nemotron-nano-8b-v1\",\"type\":\"model\"}\n95:{\"name\":\"gemma-3-27b-it\",\"type\":\"model\"}\n96:{\"name\":\"phi-4-multimodal-instruct\",\"type\":\"model\"}\n97:{\"name\":\"deepseek-r1-distill-llama-70b\",\"type\":\"model\"}\n98:{\"name\":\"cosmos-predict1-7b\",\"type\":\"model\"}\n"])</script><script>self.__next_f.push([1,"37:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"llama-3_3-nemotron-super-49b-v1\",\"displayName\":\"llama-3.3-nemotron-super-49b-v1\",\"publisher\":\"nvidia\",\"shortDescription\":\"High efficiency model with leading accuracy for reasoning, tool calling, chat, and instruction following.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llama-3_3-nemotron-super-49b-v1.jpg\",\"labels\":[\"advanced reasoning\",\"function calling\",\"instruction following\",\"math\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Participation considerations from adversely impacted groups (protected classes) in model design and testing:|None|\\n|Measures taken to mitigate against unwanted bias:|None|\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T18:56:30.758Z\",\"description\":\"$7b\",\"explainability\":\"$7c\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Generatable or Reverse engineerable personally-identifiable information?|None|\\n|Was consent obtained for any personal data used?|None Known|\\n|Personal data used to create this model?|None Known|\\n|How often is dataset reviewed?|Before Release|\\n|Is there provenance for all datasets used in training?|Yes|\\n|Does data labeling (annotation, metadata) comply with privacy laws?|Yes|\\n|Applicable NVIDIA Privacy Policy|https://www.nvidia.com/en-us/about-nvidia/privacy-policy/|\",\"safetyAndSecurity\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Model Application(s):|Chat, Instruction Following, Chatbot Development, Code Generation, Reasoning|\\n|Describe life critical application (if present):|None Known (please see referenced Known Risks in the Explainability subcard).|\\n|Use Case Restrictions:|Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.3 Community License Agreement](https://www.llama.com/llama3_3/license/). Built with Llama.|\\n|Model and Dataset Restrictions:|The Principle of least privilege (PoLP) is applied limiting access for dataset generation. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face and NGC, and may become available on cloud providers' model catalog.|\",\"updatedDate\":\"2025-03-25T21:52:27.860Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"47ecea08-d687-4ed1-8cb1-72f1891c1186\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nvidia/llama-3.3-nemotron-super-49b-v1\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_3-nemotron-super-49b-v1 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Llama 3.3 License\",\"url\":\"https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/LICENSE\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}}}},\"402\":{\"description\":\"Payment Required\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"system\\\",\\n \\\"content\\\": \\\"You are a helpful assistant.\\\"\\n },\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stop\\\": null,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"Here's a short poem on...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"What can I see at NVIDIA's GPU Technology Conference?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"system\\\",\\n \\\"content\\\": \\\"You are a helpful assistant.\\\"\\n },\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What can I see at NVIDIA's GPU Technology Conference?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stop\\\": null,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GTC conference...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n base_url = \\\"https://integrate.api.nvidia.com/v1\\\",\\n api_key = \\\"$NVIDIA_API_KEY\\\"\\n)\\n\\ncompletion = client.chat.completions.create(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n frequency_penalty=\u003c%- request.frequency_penalty %\u003e,\\n presence_penalty=\u003c%- request.presence_penalty %\u003e,\\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in completion:\\n if chunk.choices[0].delta.content is not None:\\n print(chunk.choices[0].delta.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nprint(completion.choices[0].message)\\n\u003c% } %\u003e\\n\",\"node.js\":\"import OpenAI from 'openai';\\n\\nconst openai = new OpenAI({\\n apiKey: '$NVIDIA_API_KEY',\\n baseURL: 'https://integrate.api.nvidia.com/v1',\\n})\\n\\nasync function main() {\\n const completion = await openai.chat.completions.create({\\n model: \\\"\u003c%- request.model %\u003e\\\",\\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature: \u003c%- request.temperature %\u003e,\\n top_p: \u003c%- request.top_p %\u003e,\\n max_tokens: \u003c%- request.max_tokens %\u003e,\\n frequency_penalty: \u003c%- request.frequency_penalty %\u003e,\\n presence_penalty: \u003c%- request.presence_penalty %\u003e,\\n stream: \u003c%- request.stream %\u003e,\\n })\\n \u003c% if (request.stream) { %\u003e\\n for await (const chunk of completion) {\\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\\n }\\n \u003c% } else { %\u003e\\n process.stdout.write(completion.choices[0]?.message?.content);\\n \u003c% } %\u003e\\n}\\n\\nmain();\",\"curl\":\"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature %\u003e, \\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"frequency_penalty\\\": \u003c%- request.frequency_penalty %\u003e,\\n \\\"presence_penalty\\\": \u003c%- request.presence_penalty %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e \\n }'\\n\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.3-nemotron-super-49b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"ChatCompletionRequest\":{\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"nvidia/llama-3.3-nemotron-super-49b-v1\"},\"max_tokens\":{\"type\":\"integer\",\"minimum\":1,\"title\":\"Max Tokens\",\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"default\":4096,\"maximum\":16384},\"stream\":{\"type\":\"boolean\",\"title\":\"Stream\",\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"default\":false},\"temperature\":{\"type\":\"number\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"default\":0.6},\"top_p\":{\"type\":\"number\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"default\":0.95},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\",\"examples\":[null]},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"seed\":{\"type\":\"integer\",\"maximum\":18446744073709552000,\"minimum\":0,\"title\":\"Seed\",\"description\":\"The model generates random results. Changing the input seed alone will produce a different response with similar characteristics. It is possible to reproduce results by fixing the input seed (assuming all other hyperparameters are also fixed).\",\"default\":0},\"messages\":{\"anyOf\":[{\"items\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\"},\"type\":\"array\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far.\",\"examples\":[[{\"role\":\"user\",\"content\":\"Write a limerick about the wonders of GPU computing.\"}]]}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"messages\"],\"title\":\"ChatCompletionRequest\",\"description\":\"OpenAI ChatCompletionRequest\"},\"ChatCompletionResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\",\"description\":\"A unique identifier for the completion.\"},\"object\":{\"type\":\"string\",\"title\":\"Object\",\"default\":\"chat.completion\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\",\"example\":\"nvidia/llama-3.3-nemotron-super-49b-v1\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseChoice\"},\"type\":\"array\",\"title\":\"Choices\",\"description\":\"The list of completion choices the model generated for the input prompt.\"},\"usage\":{\"$ref\":\"#/components/schemas/UsageInfo\",\"description\":\"Usage statistics for the completion request.\"}},\"type\":\"object\",\"required\":[\"model\",\"choices\",\"usage\"],\"title\":\"ChatCompletionResponse\"},\"ChatCompletionResponseChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\",\"description\":\"The index of the choice in the list of choices (always 0).\"},\"message\":{\"$ref\":\"#/components/schemas/ChatMessage\",\"description\":\"A chat completion message generated by the model.\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\",\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished\"}},\"type\":\"object\",\"required\":[\"index\",\"message\"],\"title\":\"ChatCompletionResponseChoice\"},\"ChatMessage\":{\"properties\":{\"role\":{\"type\":\"string\",\"title\":\"Role\",\"description\":\"The role of the message author.\"},\"content\":{\"type\":\"string\",\"title\":\"Content\",\"description\":\"The contents of the message.\"}},\"type\":\"object\",\"required\":[\"role\",\"content\"],\"title\":\"ChatMessage\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\",\"description\":\"Detailed information about the error.\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"PaymentRequiredError\":{\"properties\":{\"detail\":{\"type\":\"string\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"},\"UsageInfo\":{\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"title\":\"Prompt Tokens\",\"description\":\"Number of tokens in the prompt.\",\"default\":0},\"total_tokens\":{\"type\":\"integer\",\"title\":\"Total Tokens\",\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"default\":0},\"completion_tokens\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Completion Tokens\",\"description\":\"Number of tokens in the generated completion.\",\"default\":0}},\"type\":\"object\",\"title\":\"UsageInfo\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\",\"description\":\"The error message.\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\",\"description\":\"Error type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T21:52:28.660Z\",\"nvcfFunctionId\":\"a2646582-44bd-4899-bdf4-acf8c9df702c\",\"createdDate\":\"2025-03-18T18:56:31.023Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_3-nemotron-super-49b-v1\",\"termsOfUse\":\"b\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e AI Foundation Models Community License Agreement \u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/LICENSE\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e Llama 3.3 Community License Agreement \u003c/a\u003e, Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"playground\":{\"type\":\"chatWithThinking\",\"options\":{\"chat\":{\"reasoning\":{\"defaultEnabled\":true,\"systemPromptEnabled\":\"detailed thinking on\",\"systemPromptDisabled\":\"detailed thinking off\",\"toolsEnabledWithReasoning\":false,\"toolsEnabledWithoutReasoning\":true}}}},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$7d\"}]},\"artifactName\":\"llama-3_3-nemotron-super-49b-v1\"},\"config\":{\"name\":\"llama-3_3-nemotron-super-49b-v1\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"deepseek-r1\",\"displayName\":\"deepseek-r1\",\"publisher\":\"deepseek-ai\",\"shortDescription\":\"State-of-the-art, high-efficiency LLM excelling in reasoning, math, and coding.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/deepseek-r1.jpg\",\"labels\":[\"Math\",\"advanced reasoning\",\"chat\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-30T23:29:47.817Z\",\"description\":\"$7e\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-15T15:30:41.527Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"bb350840-a3f5-4a69-8438-a43f06ff0baa\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for deepseek-ai/deepseek-r1\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"MIT License\",\"url\":\"https://huggingface.co/deepseek-ai/deepseek-r1/blob/main/LICENSE\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletion\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionChunk\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Which number is larger, 9.11 or 9.8?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Which number is larger, 9.11 or 9.8?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The number 9.11 is...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"How many 'r's are in 'strawberry'?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"How many 'r's are in 'strawberry'?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"In the word strawberry...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"$7f\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"$80\",\"curl\":\"$81\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"ChatCompletion\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/Choice\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Usage\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"},\"ChatCompletionChunk\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/ChoiceChunk\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"},\"ChatRequest\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"deepseek-ai/deepseek-r1\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.6,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":4096,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"},\"Choice\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"ChoiceChunk\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"Message\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"user\",\"assistant\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"Usage\":{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-15T15:30:42.018Z\",\"nvcfFunctionId\":\"854db4e5-9be7-45a0-a730-183cadf87e50\",\"createdDate\":\"2025-01-30T23:29:48.101Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e NVIDIA Community Model License\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eMIT License\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"playground\":{\"type\":\"chatWithThinking\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$82\"}]},\"artifactName\":\"deepseek-r1\"},\"config\":{\"name\":\"deepseek-r1\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"llama-3_1-nemotron-nano-8b-v1\",\"displayName\":\"llama-3.1-nemotron-nano-8b-v1\",\"publisher\":\"nvidia\",\"shortDescription\":\"Leading reasoning and agentic AI accuracy model for PC and edge.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llama-3_3-nemotron-49b-instruct.jpg\",\"labels\":[\"advanced reasoning\",\"function calling\",\"instruction following\",\"math\"],\"attributes\":[{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Participation considerations from adversely impacted groups (protected classes) in model design and testing:|None|\\n|Measures taken to mitigate against unwanted bias:|None|\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T18:56:30.361Z\",\"description\":\"$83\",\"explainability\":\"$84\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Generatable or Reverse engineerable personally-identifiable information?|None|\\n|Was consent obtained for any personal data used?|None Known|\\n|Personal data used to create this model?|None Known|\\n|How often is dataset reviewed?|Before Release|\\n|Is there provenance for all datasets used in training?|Yes|\\n|Does data labeling (annotation, metadata) comply with privacy laws?|Yes|\\n|Applicable NVIDIA Privacy Policy|https://www.nvidia.com/en-us/about-nvidia/privacy-policy/|\",\"safetyAndSecurity\":\"|Field:|Response:|\\n|:---:|:---:|\\n|Model Application(s):|Chat, Instruction Following, Chatbot Development, Code Generation, Reasoning|\\n|Describe life critical application (if present):|None Known (please see referenced Known Risks in the Explainability subcard).|\\n|Use Case Restrictions:|Your use of this model is governed by the [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/). Additional Information: [Llama 3.1 Community License Agreement](https://www.llama.com/llama3_1/license/). Built with Llama.|\\n|Model and Dataset Restrictions:|The Principle of least privilege (PoLP) is applied limiting access for dataset generation. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face and NGC, and may become available on cloud providers' model catalog.|\",\"updatedDate\":\"2025-03-18T18:56:30.361Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"8179df65-0b97-4873-aa50-6cdf691764d0\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nvidia/llama-3.1-nemotron-nano-8b-v1\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_1-nemotron-nano-8b-v1 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Llama 3.1 License\",\"url\":\"https://github.com/meta-llama/llama-models/blob/main/License/Llama3.1.txt\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}}}},\"402\":{\"description\":\"Payment Required\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Explain how a transformer neural network works.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"system\\\",\\n \\\"content\\\": \\\"You are a helpful assistant.\\\"\\n },\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Explain how a transformer neural network works.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stop\\\": null,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"A transformer neural network is an architecture...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"Write a Python function to find the nth Fibonacci number using dynamic programming.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"system\\\",\\n \\\"content\\\": \\\"You are a helpful assistant.\\\"\\n },\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a Python function to find the nth Fibonacci number using dynamic programming.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stop\\\": null,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"Here's a Python function to...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n base_url = \\\"https://integrate.api.nvidia.com/v1\\\",\\n api_key = \\\"$NVIDIA_API_KEY\\\"\\n)\\n\\ncompletion = client.chat.completions.create(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n frequency_penalty=\u003c%- request.frequency_penalty %\u003e,\\n presence_penalty=\u003c%- request.presence_penalty %\u003e,\\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in completion:\\n if chunk.choices[0].delta.content is not None:\\n print(chunk.choices[0].delta.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nprint(completion.choices[0].message)\\n\u003c% } %\u003e\\n\",\"node.js\":\"import OpenAI from 'openai';\\n\\nconst openai = new OpenAI({\\n apiKey: '$NVIDIA_API_KEY',\\n baseURL: 'https://integrate.api.nvidia.com/v1',\\n})\\n\\nasync function main() {\\n const completion = await openai.chat.completions.create({\\n model: \\\"\u003c%- request.model %\u003e\\\",\\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature: \u003c%- request.temperature %\u003e,\\n top_p: \u003c%- request.top_p %\u003e,\\n max_tokens: \u003c%- request.max_tokens %\u003e,\\n frequency_penalty: \u003c%- request.frequency_penalty %\u003e,\\n presence_penalty: \u003c%- request.presence_penalty %\u003e,\\n stream: \u003c%- request.stream %\u003e,\\n })\\n \u003c% if (request.stream) { %\u003e\\n for await (const chunk of completion) {\\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\\n }\\n \u003c% } else { %\u003e\\n process.stdout.write(completion.choices[0]?.message?.content);\\n \u003c% } %\u003e\\n}\\n\\nmain();\",\"curl\":\"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature %\u003e, \\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"frequency_penalty\\\": \u003c%- request.frequency_penalty %\u003e,\\n \\\"presence_penalty\\\": \u003c%- request.presence_penalty %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e \\n }'\\n\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nvidia/llama-3.1-nemotron-nano-8b-v1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"ChatCompletionRequest\":{\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"nvidia/llama-3.1-nemotron-nano-8b-v1\"},\"max_tokens\":{\"type\":\"integer\",\"minimum\":1,\"title\":\"Max Tokens\",\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"default\":4096,\"maximum\":16384},\"stream\":{\"type\":\"boolean\",\"title\":\"Stream\",\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"default\":false},\"temperature\":{\"type\":\"number\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"default\":0.6},\"top_p\":{\"type\":\"number\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"default\":0.95},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\",\"examples\":[null]},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"seed\":{\"type\":\"integer\",\"maximum\":18446744073709552000,\"minimum\":0,\"title\":\"Seed\",\"description\":\"The model generates random results. Changing the input seed alone will produce a different response with similar characteristics. It is possible to reproduce results by fixing the input seed (assuming all other hyperparameters are also fixed).\",\"default\":0},\"messages\":{\"anyOf\":[{\"items\":{\"additionalProperties\":{\"type\":\"string\"},\"type\":\"object\"},\"type\":\"array\"}],\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far.\",\"examples\":[[{\"role\":\"user\",\"content\":\"Write a limerick about the wonders of GPU computing.\"}]]}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"messages\"],\"title\":\"ChatCompletionRequest\",\"description\":\"OpenAI ChatCompletionRequest\"},\"ChatCompletionResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\",\"description\":\"A unique identifier for the completion.\"},\"object\":{\"type\":\"string\",\"title\":\"Object\",\"default\":\"chat.completion\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\",\"example\":\"nvidia/llama-3.1-nemotron-nano-8b-v1\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseChoice\"},\"type\":\"array\",\"title\":\"Choices\",\"description\":\"The list of completion choices the model generated for the input prompt.\"},\"usage\":{\"$ref\":\"#/components/schemas/UsageInfo\",\"description\":\"Usage statistics for the completion request.\"}},\"type\":\"object\",\"required\":[\"model\",\"choices\",\"usage\"],\"title\":\"ChatCompletionResponse\"},\"ChatCompletionResponseChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\",\"description\":\"The index of the choice in the list of choices (always 0).\"},\"message\":{\"$ref\":\"#/components/schemas/ChatMessage\",\"description\":\"A chat completion message generated by the model.\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\",\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished\"}},\"type\":\"object\",\"required\":[\"index\",\"message\"],\"title\":\"ChatCompletionResponseChoice\"},\"ChatMessage\":{\"properties\":{\"role\":{\"type\":\"string\",\"title\":\"Role\",\"description\":\"The role of the message author.\"},\"content\":{\"type\":\"string\",\"title\":\"Content\",\"description\":\"The contents of the message.\"}},\"type\":\"object\",\"required\":[\"role\",\"content\"],\"title\":\"ChatMessage\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\",\"description\":\"Detailed information about the error.\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"PaymentRequiredError\":{\"properties\":{\"detail\":{\"type\":\"string\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"},\"UsageInfo\":{\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"title\":\"Prompt Tokens\",\"description\":\"Number of tokens in the prompt.\",\"default\":0},\"total_tokens\":{\"type\":\"integer\",\"title\":\"Total Tokens\",\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"default\":0},\"completion_tokens\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Completion Tokens\",\"description\":\"Number of tokens in the generated completion.\",\"default\":0}},\"type\":\"object\",\"title\":\"UsageInfo\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\",\"description\":\"The error message.\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\",\"description\":\"Error type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"nvcfFunctionId\":\"07587d89-2056-4eb8-a448-4b05463d2478\",\"createdDate\":\"2025-03-18T18:56:30.714Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_1-nemotron-nano-8b-v1\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e AI Foundation Models Community License Agreement \u003c/a\u003e. ADDITIONAL INFORMATION: Llama 3.1 Community License Agreement, Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"playground\":{\"type\":\"chatWithThinking\",\"options\":{\"chat\":{\"reasoning\":{\"defaultEnabled\":false,\"systemPromptEnabled\":\"detailed thinking on\",\"systemPromptDisabled\":\"detailed thinking off\",\"toolsEnabledWithReasoning\":false,\"toolsEnabledWithoutReasoning\":true}}}}},\"artifactName\":\"llama-3_1-nemotron-nano-8b-v1\"},\"config\":{\"name\":\"llama-3_1-nemotron-nano-8b-v1\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"gemma-3-27b-it\",\"displayName\":\"gemma-3-27b-it\",\"publisher\":\"google\",\"shortDescription\":\"Cutting-edge open multimodal model exceling in high-quality reasoning from images.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/gemma-3-27b-it.jpg\",\"labels\":[\"Language Generation\",\"Vision Assistant\",\"Visual Question Answering\",\"Image-to-Text\"],\"attributes\":[{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-13T17:31:09.671Z\",\"description\":\"$85\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-14T19:44:50.785Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"2dbfe848-9253-4fd6-b0d6-b70e823deb1b\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for google/gemma-3-27b-it\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/google-gemma-3-27b-it for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"Gemma Terms of Use\",\"url\":\"https://ai.google.dev/gemma/terms\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1\"}],\"tags\":[{\"name\":\"Multimodal API\",\"description\":\"This API performs inference using multi-modal large language models (MLLMs).\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"tags\":[\"Multimodal API\"],\"summary\":\"Request response from the model\",\"description\":\"Invokes inference using the model chat parameters. If uploading large images, this POST should be used in conjunction with the NVCF API which allows for the upload of large assets. \\nYou can find details on how to use NVCF Asset APIs here: https://docs.api.nvidia.com/cloud-functions/reference/createasset\",\"operationId\":\"invokeFunction\",\"parameters\":[{\"in\":\"header\",\"name\":\"NVCF-INPUT-ASSET-REFERENCES\",\"schema\":{\"type\":\"string\",\"maxLength\":370,\"format\":\"uuid\"},\"required\":false,\"description\":\"String of asset IDs separated by commas. Data is uploaded to AWS S3 using NVCF Asset APIs and associated with these asset IDs.If the size of an image is more than 180KB, it needs to be uploaded to a presigned S3 URL bucket. The presigned URL allows for secure and temporary access to the S3 bucket for uploading the image. Once the asset is requested, an asset ID is generated for it. Please include this asset ID in this header and to use the uploaded image in a prompt, you need to refer to it using the following format: `\u003cimg src=\\\"data:image/png;asset_id,{asset_id}\\\" /\u003e`.\"}],\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionStreamResponse\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"image_captioning\",\"input\":{\"text\":\"What is shown in this image?\",\"images\":[\"https://assets.ngc.nvidia.com/products/api-catalog/gemma3/traffic_sign.jpg\"]},\"output\":{\"text\":\"**A red stop sign:** Prominently displayed in the foreground...\"}},{\"name\":\"visual_qa\",\"input\":{\"text\":\"Where is the cow standing?\",\"images\":[\"https://assets.ngc.nvidia.com/products/api-catalog/gemma3/cow_on_the_beach.jpg\"]},\"output\":{\"text\":\"On a beach.\"}}],\"templates\":[{\"title\":\"Default\",\"requestEjs\":{\"python\":\"$86\",\"node.js\":\"$87\",\"curl\":\"stream=\u003c%- request.stream %\u003e\\n\\nif [ \\\"$stream\\\" = true ]; then\\n accept_header='Accept: text/event-stream'\\nelse\\n accept_header='Accept: application/json'\\nfi\\n\u003c% content = \\\"What is in this image? \u003cimg src=\\\\\\\\\\\\\\\"data:image/png;base64,'\\\\\\\"\\\\$image_b64\\\\\\\"'\\\\\\\\\\\\\\\" /\u003e\\\" %\u003e\\nimage_b64=$( base64 image.png )\\n\\necho '{\\n \\\"model\\\": \\\"\u003c%- request.model %\u003e\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"\u003c%- content %\u003e\\\"\\n }\\n ],\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature.toFixed(2) %\u003e,\\n \\\"top_p\\\": \u003c%- request.top_p.toFixed(2) %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e\\n}' \u003e payload.json\\n\\ncurl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"$accept_header\\\" \\\\\\n -d @payload.json\\n\"},\"response\":\"{\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"model\\\": \\\"google/gemma-3-27b-it\\\",\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"...\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}},\"/status/{requestId}\":{\"get\":{\"tags\":[\"Multimodal API\"],\"summary\":\"Gets the result of an earlier function invocation request that returned a status of 202.\",\"operationId\":\"getFunctionInvocationResult\",\"parameters\":[{\"name\":\"requestId\",\"in\":\"path\",\"description\":\"requestId to poll results\",\"required\":true,\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}}],\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"422\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"ChatCompletionResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"object\":{\"type\":\"string\",\"enum\":[\"chat.completion\"],\"const\":\"chat.completion\",\"title\":\"Object\",\"default\":\"chat.completion\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseChoice\"},\"type\":\"array\",\"title\":\"Choices\"},\"usage\":{\"$ref\":\"#/components/schemas/UsageInfo\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"model\",\"choices\",\"usage\"],\"title\":\"ChatCompletionResponse\"},\"ChatCompletionResponseChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\"},\"message\":{\"$ref\":\"#/components/schemas/ChatMessage\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\"},\"stop_reason\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop Reason\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"index\",\"message\"],\"title\":\"ChatCompletionResponseChoice\"},\"ChatCompletionResponseStreamChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\"},\"delta\":{\"$ref\":\"#/components/schemas/DeltaMessage\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\"},\"stop_reason\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop Reason\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"index\",\"delta\"],\"title\":\"ChatCompletionResponseStreamChoice\"},\"ChatCompletionStreamResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"object\":{\"type\":\"string\",\"enum\":[\"chat.completion.chunk\"],\"const\":\"chat.completion.chunk\",\"title\":\"Object\",\"default\":\"chat.completion.chunk\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseStreamChoice\"},\"type\":\"array\",\"title\":\"Choices\"},\"usage\":{\"anyOf\":[{\"$ref\":\"#/components/schemas/UsageInfo\"},{\"type\":\"null\"}]}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"model\",\"choices\"],\"title\":\"ChatCompletionStreamResponse\"},\"ChatMessage\":{\"properties\":{\"role\":{\"type\":\"string\",\"title\":\"Role\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Content\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"role\"],\"title\":\"ChatMessage\"},\"DeltaMessage\":{\"properties\":{\"role\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Role\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Content\"}},\"additionalProperties\":false,\"type\":\"object\",\"title\":\"DeltaMessage\"},\"ErrorResponse\":{\"properties\":{\"object\":{\"type\":\"string\",\"title\":\"Object\",\"default\":\"error\"},\"message\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Type\"},\"param\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Param\"},\"code\":{\"type\":\"integer\",\"title\":\"Code\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"message\",\"type\",\"code\"],\"title\":\"ErrorResponse\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"ImageURL\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\"}},\"type\":\"object\",\"required\":[\"url\"],\"title\":\"ImageURL\"},\"NIMLLMChatCompletionMessage\":{\"properties\":{\"role\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Role\"}],\"description\":\"The role of the message's author.\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"items\":{\"anyOf\":[{\"$ref\":\"#/components/schemas/NIMVLMChatCompletionContentPartImage\"},{\"$ref\":\"#/components/schemas/NIMVLMChatCompletionContentPartText\"}]},\"type\":\"array\"}],\"title\":\"Content\",\"description\":\"The contents of the message.\\n \u003cbr\u003eTo pass images (only with role=`user`):\\n \u003cbr\u003e - When content is a string, image can be passed together with the text with `img` HTML tags that wraps \\n an image URL (`\u003cimg src=\\\"{url}\\\" /\u003e`), \\n base64 encoded image data (`\u003cimg src=\\\"data:image/{format};base64,{base64encodedimage}\\\" /\u003e`), \\n or an NVCF asset ID (`\u003cimg src=\\\"data:image/{format};asset_id,{asset_id}\\\" /\u003e`) \\n when the container is hosted in NVCF and the payload exceeds 200KB.\\n \u003cbr\u003e - When content is a list of objects, images can be passed as objects with type=`image_url`.\\n \u003cbr\u003e - In both cases, images can be PNG, JPG or JPEG.\\n \u003cbr\u003eFor `system` and `assistant` roles, the object list format is not supported.\\n \"}},\"type\":\"object\",\"required\":[\"role\",\"content\"],\"title\":\"NIMLLMChatCompletionMessage\"},\"NIMLLMChatCompletionRequest\":{\"properties\":{\"messages\":{\"items\":{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionMessage\"},\"type\":\"array\",\"minItems\":1,\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far.\"},\"model\":{\"type\":\"string\",\"title\":\"Model\",\"description\":\"The model to use.\",\"default\":\"google/gemma-3-27b-it\"},\"max_tokens\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":4096,\"minimum\":1},{\"type\":\"null\"}],\"title\":\"Max Tokens\",\"description\":\"The maximum number of tokens that can be generated.\",\"default\":512},\"stream\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"Stream\",\"description\":\"If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format) as they become available, with the stream terminated by a `data: [DONE]`\",\"default\":false},\"temperature\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":2,\"minimum\":0},{\"type\":\"null\"}],\"title\":\"Temperature\",\"description\":\"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.\",\"default\":0.2},\"top_p\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":1,\"exclusiveMinimum\":0},{\"type\":\"null\"}],\"title\":\"Top P\",\"description\":\"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.\",\"default\":0.7}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"messages\",\"model\"],\"title\":\"NIMLLMChatCompletionRequest\"},\"NIMVLMChatCompletionContentPartImage\":{\"properties\":{\"image_url\":{\"$ref\":\"#/components/schemas/ImageURL\"},\"type\":{\"type\":\"string\",\"enum\":[\"image_url\"],\"const\":\"image_url\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"image_url\",\"type\"],\"title\":\"NIMVLMChatCompletionContentPartImage\"},\"NIMVLMChatCompletionContentPartText\":{\"properties\":{\"text\":{\"type\":\"string\",\"title\":\"Text\"},\"type\":{\"type\":\"string\",\"enum\":[\"text\"],\"const\":\"text\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"text\",\"type\"],\"title\":\"NIMVLMChatCompletionContentPartText\"},\"Role\":{\"type\":\"string\",\"enum\":[\"assistant\",\"user\"],\"title\":\"Role\"},\"UsageInfo\":{\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"title\":\"Prompt Tokens\",\"default\":0},\"total_tokens\":{\"type\":\"integer\",\"title\":\"Total Tokens\",\"default\":0},\"completion_tokens\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Completion Tokens\",\"default\":0}},\"additionalProperties\":false,\"type\":\"object\",\"title\":\"UsageInfo\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-14T19:44:51.315Z\",\"nvcfFunctionId\":\"b86482fe-8e57-4f0a-8efd-7a45a985377e\",\"createdDate\":\"2025-03-13T17:31:09.933Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/google-gemma-3-27b-it\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Community Model License\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://ai.google.dev/gemma/terms\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eGemma Terms of Use\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"playground\":{\"type\":\"chatWithImages\",\"options\":{\"image\":{\"singleTurn\":false,\"limit\":3}}}},\"artifactName\":\"gemma-3-27b-it\"},\"config\":{\"name\":\"gemma-3-27b-it\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"phi-4-multimodal-instruct\",\"displayName\":\"phi-4-multimodal-instruct\",\"publisher\":\"microsoft\",\"shortDescription\":\"Cutting-edge open multimodal model exceling in high-quality reasoning from image and audio inputs.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/phi-4-multimodal-instruct.jpg\",\"labels\":[\"Chart and Table Understanding\",\"Language Generation\",\"Speech Recognition\",\"Visual QA\",\"Image-to-Text\"],\"attributes\":[{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-02-26T21:16:43.734Z\",\"description\":\"$88\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-26T21:29:52.872Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"feae5394-8ae3-4218-a3fd-20fe10cc8f66\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for microsoft/phi-4-multimodal-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/microsoft-phi-4-multimodal-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"MIT\",\"url\":\"https://opensource.org/license/mit\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1\"}],\"tags\":[{\"name\":\"Multimodal API\",\"description\":\"This API performs inference using multi-modal large language models (MLLMs).\"}],\"paths\":{\"/microsoft/phi-4-multimodal-instruct\":{\"post\":{\"tags\":[\"Multimodal API\"],\"summary\":\"Request response from the model\",\"description\":\"Invokes inference using the model chat parameters. If uploading large images or audios, this POST should be used in conjunction with the NVCF API which allows for the upload of large assets. \\nYou can find details on how to use NVCF Asset APIs here: https://docs.api.nvidia.com/cloud-functions/reference/createasset\",\"operationId\":\"invokeFunction\",\"parameters\":[{\"in\":\"header\",\"name\":\"NVCF-INPUT-ASSET-REFERENCES\",\"schema\":{\"type\":\"string\",\"maxLength\":370,\"format\":\"uuid\"},\"required\":false,\"description\":\"String of asset IDs separated by commas. Data is uploaded to AWS S3 using NVCF Asset APIs and associated with these asset IDs. If the size of an image or audio is more than 180KB, it needs to be uploaded to a presigned S3 URL bucket. The presigned URL allows for secure and temporary access to the S3 bucket for uploading the image or audio. Once the asset is requested, an asset ID is generated for it. Please include this asset ID in this header. Insert an uploaded image in a prompt using the following format: `\u003cimg src=\\\"data:image/png;asset_id,{asset_id}\\\" /\u003e`. Insert an uploaded audio in a prompt using the following format: `\u003caudio src=\\\"data:audio/wav;asset_id,{asset_id}\\\" /\u003e`\"}],\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionStreamResponse\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"image_captioning\",\"input\":{\"text\":\"Describe this image.\",\"images\":[\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/traffic_sign.jpg\"]}},{\"name\":\"multi_image_comparison\",\"input\":{\"text\":\"What are common about these two images?\",\"images\":[\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/cats.jpeg\",\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/dogs.jpeg\"]}},{\"name\":\"speech_recognition\",\"input\":{\"text\":\"Transcribe the spoken content.\",\"audios\":[\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/what_is_the_traffic_sign_in_the_image.wav\"]}},{\"name\":\"audio_visual_understanding\",\"input\":{\"text\":\"\",\"audios\":[\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/what_is_the_traffic_sign_in_the_image.wav\"],\"images\":[\"https://assets.ngc.nvidia.com/products/api-catalog/phi-4-mini-mm/traffic_sign.jpg\"]}}],\"templates\":[{\"title\":\"Default\",\"requestEjs\":{\"python\":\"$89\",\"node.js\":\"$8a\",\"curl\":\"stream=\u003c%- request.stream %\u003e\\n\\nif [ \\\"$stream\\\" = true ]; then\\n accept_header='Accept: text/event-stream'\\nelse\\n accept_header='Accept: application/json'\\nfi\\n\u003c% content = \\\"Answer the spoken query about the image.\u003cimg src=\\\\\\\\\\\\\\\"data:image/png;base64,'\\\\\\\"\\\\$image_b64\\\\\\\"'\\\\\\\\\\\\\\\" /\u003e\u003caudio src=\\\\\\\\\\\\\\\"data:audio/wav;base64,'\\\\\\\"\\\\$audio_b64\\\\\\\"'\\\\\\\\\\\\\\\" /\u003e\\\" %\u003e\\nimage_b64=$( base64 image.png )\\naudio_b64=$( base64 audio.wav )\\n\\necho '{\\n \\\"model\\\": \\\"\u003c%- request.model %\u003e\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"\u003c%- content %\u003e\\\"\\n }\\n ],\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature.toFixed(2) %\u003e,\\n \\\"top_p\\\": \u003c%- request.top_p.toFixed(2) %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e\\n}' \u003e payload.json\\n\\ncurl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"$accept_header\\\" \\\\\\n -d @payload.json\\n\"},\"response\":\"{\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"model\\\": \\\"microsoft/phi-4-multimodal-instruct\\\",\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"...\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}},\"/status/{requestId}\":{\"get\":{\"tags\":[\"Multimodal API\"],\"summary\":\"Gets the result of an earlier function invocation request that returned a status of 202.\",\"operationId\":\"getFunctionInvocationResult\",\"parameters\":[{\"name\":\"requestId\",\"in\":\"path\",\"description\":\"requestId to poll results\",\"required\":true,\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}}],\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionResponse\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\",\"format\":\"^[a-zA-Z-]{1,64}$\",\"maxLength\":64}}}},\"422\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ErrorResponse\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/e598bfc1-b058-41af-869d-556d3c7e1b48\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"ChatCompletionResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"object\":{\"type\":\"string\",\"enum\":[\"chat.completion\"],\"const\":\"chat.completion\",\"title\":\"Object\",\"default\":\"chat.completion\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseChoice\"},\"type\":\"array\",\"title\":\"Choices\"},\"usage\":{\"$ref\":\"#/components/schemas/UsageInfo\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"model\",\"choices\",\"usage\"],\"title\":\"ChatCompletionResponse\"},\"ChatCompletionResponseChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\"},\"message\":{\"$ref\":\"#/components/schemas/ChatMessage\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\"},\"stop_reason\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop Reason\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"index\",\"message\"],\"title\":\"ChatCompletionResponseChoice\"},\"ChatCompletionResponseStreamChoice\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\"},\"delta\":{\"$ref\":\"#/components/schemas/DeltaMessage\"},\"finish_reason\":{\"anyOf\":[{\"type\":\"string\",\"enum\":[\"stop\",\"length\"]},{\"type\":\"null\"}],\"title\":\"Finish Reason\"},\"stop_reason\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop Reason\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"index\",\"delta\"],\"title\":\"ChatCompletionResponseStreamChoice\"},\"ChatCompletionStreamResponse\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"object\":{\"type\":\"string\",\"enum\":[\"chat.completion.chunk\"],\"const\":\"chat.completion.chunk\",\"title\":\"Object\",\"default\":\"chat.completion.chunk\"},\"created\":{\"type\":\"integer\",\"title\":\"Created\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"choices\":{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionResponseStreamChoice\"},\"type\":\"array\",\"title\":\"Choices\"},\"usage\":{\"anyOf\":[{\"$ref\":\"#/components/schemas/UsageInfo\"},{\"type\":\"null\"}]}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"model\",\"choices\"],\"title\":\"ChatCompletionStreamResponse\"},\"ChatMessage\":{\"properties\":{\"role\":{\"type\":\"string\",\"title\":\"Role\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Content\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"role\"],\"title\":\"ChatMessage\"},\"DeltaMessage\":{\"properties\":{\"role\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Role\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Content\"}},\"additionalProperties\":false,\"type\":\"object\",\"title\":\"DeltaMessage\"},\"ErrorResponse\":{\"properties\":{\"object\":{\"type\":\"string\",\"title\":\"Object\",\"default\":\"error\"},\"message\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Type\"},\"param\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Param\"},\"code\":{\"type\":\"integer\",\"title\":\"Code\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"message\",\"type\",\"code\"],\"title\":\"ErrorResponse\"},\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"AudioURL\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\"}},\"type\":\"object\",\"required\":[\"url\"],\"title\":\"AudioURL\"},\"InputAudio\":{\"properties\":{\"data\":{\"type\":\"string\",\"title\":\"Data\"},\"format\":{\"type\":\"string\",\"enum\":[\"wav\",\"mp3\"],\"title\":\"Format\"}},\"type\":\"object\",\"required\":[\"data\",\"format\"],\"title\":\"InputAudio\"},\"ImageURL\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\"}},\"type\":\"object\",\"required\":[\"url\"],\"title\":\"ImageURL\"},\"NIMLLMChatCompletionMessage\":{\"properties\":{\"role\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Role\"}],\"description\":\"The role of the message's author.\"},\"content\":{\"anyOf\":[{\"type\":\"string\"},{\"items\":{\"anyOf\":[{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionContentPartAudio\"},{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionContentPartImage\"},{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionContentPartText\"}]},\"type\":\"array\"}],\"title\":\"Content\",\"description\":\"The contents of the message.\\n \u003cbr\u003eTo pass images or audios (only with role=`user`):\\n \u003cbr\u003e- When content is a string, image or audio can be passed together with the text with HTML-style tags that wraps an image or audio URL (`\u003cimg src=\\\"{url}\\\" /\u003e` or `\u003caudio src=\\\"{url}\\\" /\u003e`), base64 encoded image or audio data (`\u003cimg src=\\\"data:image/{format};base64,{base64encodedimage}\\\" /\u003e` or `\u003caudio src=\\\"data:audio/{format};base64,{base64encodedaudio}\\\" /\u003e`), or an NVCF asset ID (`\u003cimg src=\\\"data:image/{format};asset_id,{asset_id}\\\" /\u003e` or `\u003caudio src=\\\"data:audio/{format};asset_id,{asset_id}\\\" /\u003e`) when the container is hosted in NVCF and the payload exceeds 200KB. \u003cbr\u003e- When content is a list of objects, images can be passed as objects with type=`image_url`, and audios can be passed as objects with type=`audio_url` or `input_audio`.\\n \u003cbr\u003e- In both cases, images can be PNG, JPG or JPEG, and audios can be WAV or MP3.\\n \u003cbr\u003eFor `system` and `assistant` roles, the object list format is not supported.\"}},\"type\":\"object\",\"required\":[\"role\",\"content\"],\"title\":\"NIMLLMChatCompletionMessage\"},\"NIMLLMChatCompletionRequest\":{\"properties\":{\"messages\":{\"items\":{\"$ref\":\"#/components/schemas/NIMLLMChatCompletionMessage\"},\"type\":\"array\",\"minItems\":1,\"title\":\"Messages\",\"description\":\"A list of messages comprising the conversation so far.\"},\"model\":{\"type\":\"string\",\"title\":\"Model\",\"description\":\"The model to use.\",\"default\":\"microsoft/phi-4-multimodal-instruct\"},\"frequency_penalty\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":2,\"minimum\":-2},{\"type\":\"null\"}],\"title\":\"Frequency Penalty\",\"description\":\"Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.\",\"default\":0},\"max_tokens\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":8192,\"minimum\":1},{\"type\":\"null\"}],\"title\":\"Max Tokens\",\"description\":\"The maximum number of tokens that can be generated.\",\"default\":512},\"presence_penalty\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":2,\"minimum\":-2},{\"type\":\"null\"}],\"title\":\"Presence Penalty\",\"description\":\"Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.\",\"default\":0},\"seed\":{\"anyOf\":[{\"type\":\"integer\",\"maximum\":9223372036854776000,\"minimum\":-9223372036854776000},{\"type\":\"null\"}],\"title\":\"Seed\",\"description\":\"Changing the seed will produce a different response with similar characteristics. Fixing the seed will reproduce the same results if all other parameters are also kept constant.\"},\"stop\":{\"anyOf\":[{\"type\":\"string\"},{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"Sequences where the API will stop generating further tokens.\"},\"stream\":{\"anyOf\":[{\"type\":\"boolean\"},{\"type\":\"null\"}],\"title\":\"Stream\",\"description\":\"If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only [server-sent events](https://developer.mozilla.org/en-US/docs/Web/API/Server-sent_events/Using_server-sent_events#Event_stream_format) as they become available, with the stream terminated by a `data: [DONE]`\",\"default\":false},\"temperature\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":2,\"minimum\":0},{\"type\":\"null\"}],\"title\":\"Temperature\",\"description\":\"What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.\",\"default\":0.1},\"top_p\":{\"anyOf\":[{\"type\":\"number\",\"maximum\":1,\"exclusiveMinimum\":0},{\"type\":\"null\"}],\"title\":\"Top P\",\"description\":\"An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or `temperature` but not both.\",\"default\":0.7}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"messages\",\"model\"],\"title\":\"NIMLLMChatCompletionRequest\"},\"NIMLLMChatCompletionContentPartAudio\":{\"properties\":{\"audio_url\":{\"$ref\":\"#/components/schemas/AudioURL\"},\"type\":{\"type\":\"string\",\"enum\":[\"audio_url\"],\"const\":\"audio_url\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"audio_url\",\"type\"],\"title\":\"NIMLLMChatCompletionContentPartAudio\"},\"NIMLLMChatCompletionContentPartInputAudio\":{\"properties\":{\"input_audio\":{\"$ref\":\"#/components/schemas/InputAudio\"},\"type\":{\"type\":\"string\",\"const\":\"input_audio\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"input_audio\",\"type\"],\"title\":\"NIMLLMChatCompletionContentPartInputAudio\"},\"NIMLLMChatCompletionContentPartImage\":{\"properties\":{\"image_url\":{\"$ref\":\"#/components/schemas/ImageURL\"},\"type\":{\"type\":\"string\",\"enum\":[\"image_url\"],\"const\":\"image_url\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"image_url\",\"type\"],\"title\":\"NIMLLMChatCompletionContentPartImage\"},\"NIMLLMChatCompletionContentPartText\":{\"properties\":{\"text\":{\"type\":\"string\",\"title\":\"Text\"},\"type\":{\"type\":\"string\",\"enum\":[\"text\"],\"const\":\"text\",\"title\":\"Type\"}},\"type\":\"object\",\"required\":[\"text\",\"type\"],\"title\":\"NIMLLMChatCompletionContentPartText\"},\"Role\":{\"type\":\"string\",\"enum\":[\"assistant\",\"user\"],\"title\":\"Role\"},\"UsageInfo\":{\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"title\":\"Prompt Tokens\",\"default\":0},\"total_tokens\":{\"type\":\"integer\",\"title\":\"Total Tokens\",\"default\":0},\"completion_tokens\":{\"anyOf\":[{\"type\":\"integer\"},{\"type\":\"null\"}],\"title\":\"Completion Tokens\",\"default\":0}},\"additionalProperties\":false,\"type\":\"object\",\"title\":\"UsageInfo\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-02-26T21:29:53.631Z\",\"nvcfFunctionId\":\"3afd112d-ad66-4d29-81b6-e7d6f8225e07\",\"createdDate\":\"2025-02-26T21:16:44.266Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/microsoft-phi-4-multimodal-instruct\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; the use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Community Model License\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://huggingface.co/microsoft/Phi-4-multimodal-instruct/blob/main/LICENSE\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eMIT License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"playground\":{\"type\":\"chatWithFiles\",\"options\":{\"audio\":{\"limit\":1},\"image\":{\"singleTurn\":false,\"limit\":3}}}},\"artifactName\":\"phi-4-multimodal-instruct\"},\"config\":{\"name\":\"phi-4-multimodal-instruct\",\"type\":\"model\"}},{\"endpoint\":null,\"spec\":null,\"config\":{\"name\":\"deepseek-r1-distill-llama-70b\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"cosmos-predict1-7b\",\"displayName\":\"cosmos-predict1-7b\",\"publisher\":\"nvidia\",\"shortDescription\":\"Generates physics-aware video world states from text and image prompts for physical AI development.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/cosmos-predict1-7b.jpg\",\"labels\":[\"Physical AI\",\"image-to-world\",\"robotics\",\"text-to-world\",\"Synthetic Data Generation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"Field | Response\\n:---------------------------------------------------------------------------------------------------|:---------------\\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None\\nMeasures taken to mitigate against unwanted bias: | None\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-18T19:09:32.918Z\",\"description\":\"$8b\",\"explainability\":\"$8c\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$8d\",\"safetyAndSecurity\":\"Field | Response\\n:---------------------------------------------------|:----------------------------------\\nModel Application(s): | World Generation\\nDescribe the life critical impact (if present). | None Known\\nUse Case Restrictions: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\\nModel and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog.\",\"updatedDate\":\"2025-03-18T19:58:04.974Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"b7daa55b-4865-402a-98ef-11f9f86a4873\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim for more details.\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"license\":{\"name\":\"TODO\",\"url\":\"https://todo\"},\"title\":\"NVIDIA NIM API for nvidia/cosmos-1.0-diffusion-7b-text2world\",\"version\":\"1.0.0\"},\"servers\":[{\"url\":\"https://ai.api.nvidia.com/v1/\"}],\"paths\":{\"v1/infer\":{\"post\":{\"summary\":\" Infer\",\"operationId\":\"_infer_infer_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Text2WorldRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Text2WorldResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"templates\":[{\"requestEjs\":{\"curl\":\"$8e\",\"node.js\":\"$8f\",\"python\":\"$90\"}}]}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"Text2WorldRequest\":{\"additionalProperties\":false,\"properties\":{\"prompt\":{\"description\":\"The prompt to use for generation. The prompt describes what should be present in the output generation.\",\"title\":\"Prompt\",\"type\":\"string\"},\"seed\":{\"anyOf\":[{\"exclusiveMaximum\":4294967296,\"minimum\":0,\"type\":\"integer\"},{\"type\":\"null\"}],\"default\":0,\"description\":\"The seed which governs generation. Changing the seed with other inputs fixed results in different outputs.\",\"title\":\"Seed\"}},\"required\":[\"prompt\"],\"title\":\"Text2WorldRequest\",\"type\":\"object\"},\"Text2WorldResponse\":{\"properties\":{\"asset_url\":{\"description\":\"The URL of the generated video (mp4)\",\"examples\":[\"https://api.nvcf.nvidia.com/v1/assets/12345\"],\"title\":\"Asset Url\",\"type\":\"string\"}},\"required\":[\"asset_url\"],\"title\":\"Text2WorldResponse\",\"type\":\"object\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}},\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}}},\"security\":[{\"Token\":[]}]},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-18T19:58:05.525Z\",\"nvcfFunctionId\":\"01327741-a1cb-4bdb-a31e-5391c8ca48c2\",\"createdDate\":\"2025-03-18T19:09:33.259Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-cosmos-1_0-diffusion-7b\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS:\u003c/b\u003e This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA API Trial Terms of Service\u003c/u\u003e\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e\u003cu\u003eNVIDIA Open Model License Agreement\u003c/u\u003e\u003c/a\u003e (v. December 20, 2024).\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Download and Post-Train\",\"url\":\"$undefined\"},\"fineTuneModal\":{\"ngc\":\"https://catalog.ngc.nvidia.com/orgs/nvidia/teams/cosmos/collections/cosmos\",\"huggingFace\":\"https://huggingface.co/collections/nvidia/cosmos-6751e884dc10e013a0a0d8e6\",\"nemo\":\"https://github.com/NVIDIA/Cosmos\"},\"maxRequests\":20,\"systemCard\":\"$91\"},\"artifactName\":\"cosmos-predict1-7b\"},\"config\":{\"name\":\"cosmos-predict1-7b\",\"type\":\"model\"}}],\"items\":[\"$92\",\"$93\",\"$94\",\"$95\",\"$96\",\"$97\",\"$98\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"Most Popular Models\"}],[\"$\",\"$L26\",null,{\"href\":\"/models\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"The leading open models built by the community, optimized and accelerated by NVIDIA's enterprise-ready inference runtime.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"99:T1544,"])</script><script>self.__next_f.push([1,"**Model Overview**\n\n## Description:\n\nDeepSeek-R1-Distill-Llama-8B is a distilled version of the DeepSeek-R1 series, built upon the Llama3.1-8B-Instruct architecture. This model is designed to deliver efficient performance for reasoning, math, and code tasks while maintaining high accuracy. By distilling knowledge from the larger DeepSeek-R1 model, it provides state-of-the-art performance with reduced computational requirements.\n\nThis model is ready for both research and commercial use.\nFor more details, visit the [DeepSeek website](https://www.deepseek.com/).\n\n## Third-Party Community Consideration\n\nThis model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA [DeepSeek-R1 Model Card](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B).\n\n### License/Terms of Use\n\nGOVERNING TERMS: The NIM container is governed by the [NVIDIA Software License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/) and [Product-Specific Terms for AI Products](https://www.nvidia.com/en-us/agreements/enterprise-software/product-specific-terms-for-ai-products/); and the use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). Additional Information: [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md); [Meta Llama 3.1 Community License Agreement](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct/blob/main/LICENSE). Built with Llama.\n\n## References:\n\n- [DeepSeek GitHub Repository](https://github.com/deepseek-ai/DeepSeek-V3)\n- [DeepSeek-R1 Paper](https://arxiv.org/abs/2501.12948)\n- [Hugging Face Model Card for DeepSeek-R1-Distill-Llama-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B)\n\n## Model Architecture:\n\n**Architecture Type:** Distilled version of Mixture of Experts (MoE) \u003cbr\u003e\n**Base Model:** Llama3.1-8B-Instruct\n\n## Input:\n\n**Input Type(s):** Text \u003cbr\u003e\n**Input Format(s):** String \u003cbr\u003e\n**Input Parameters:** (1D) \u003cbr\u003e\n**Other Properties Related to Input:** \u003cbr\u003e\nDeepSeek recommends adhering to the following configurations when utilizing the DeepSeek-R1 series models, including benchmarking, to achieve the expected performance:\n\n1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.\n2. **Avoid adding a system prompt; all instructions should be contained within the user prompt.**\n3. For mathematical problems, it is advisable to include a directive in your prompt such as: \"Please reason step by step, and put your final answer within \\boxed{}.\"\n4. When evaluating model performance, it is recommended to conduct multiple tests and average the results.\n\n## Output:\n\n**Output Type(s):** Text \u003cbr\u003e\n**Output Format:** String \u003cbr\u003e\n**Output Parameters:** (1D) \u003cbr\u003e\n\n## Software Integration:\n\n**Runtime Engine(s):** TensorRT-LLM \u003cbr\u003e\n**Supported Hardware Microarchitecture Compatibility:** NVIDIA's Ampere, NVIDIA Blackwell, NVIDIA Jetson, NVIDIA Hopper, NVIDIA Lovelace, NVIDIA Pascal, NVIDIA Turing, and NVIDIA Volta architectures \u003cbr\u003e\n**[Preferred/Supported] Operating System(s):** Linux\n\n## Model Version(s):\n\nDeepSeek-R1-Distill-Llama-8B\n\n# Training, Testing, and Evaluation Datasets:\n\n## Training Dataset:\n\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n## Testing Dataset:\n\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n## Evaluation Dataset:\n\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n## Inference:\n\n**Engine:** TensorRT-LLM \u003cbr\u003e\n**Test Hardware:** NVIDIA Hopper\n\n## Ethical Considerations:\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Model Limitations:\n\nThe DeepSeek-R1-Distill model may struggle with open-ended or complex tasks, such as mathematical problems, if a directive is not included in the prompt to reason step by step and put the final answer within a boxed notation. Additionally, the model may face challenges with benchmarks requiring sampling if the temperature, top-p value, and number of responses per query are not set correctly.\n\nThe base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive."])</script><script>self.__next_f.push([1,"9a:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\n9b:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n "])</script><script>self.__next_f.push([1," messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();9c:T689,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%-"])</script><script>self.__next_f.push([1," JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003e9d:T4a6,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/deepseek-ai/deepseek-r1-distill-llama-8b:1.5.2\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"deepseek-ai/deepseek-r1-distill-llama-8b\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Which number is larger, 9.11 or 9.8?\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).9e:Tfc5,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/support-matrix.html#llama-3-2-nv-embedqa-1b-v2))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Experience via App\n\n| [ChatRTX](https://www.nvidia.com/en-us/ai-on-rtx/chatrtx/) | [AnythingLLM](https://docs.anythingllm.com/nvidia-nims/introduction) | [AI Toolkit for VSCode](https://aka.ms/aitoolkit)|\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench -u root\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod -R a+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/deepseek-ai/deepseek-r1-distill-llama-8b:1.8.0-RTX\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"deepseek-ai/deepseek-r1-distill-llama-8b\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Which number is larger, 9.11 or 9.8?\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html)\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\n"])</script><script>self.__next_f.push([1,"9f:Tbf1,"])</script><script>self.__next_f.push([1,"# Model Overview\n\n## Description:\nMistral-NeMo is a Large Language Model (LLM) composed of 12B parameters. This model leads accuracy on popular benchmarks across common sense reasoning, coding, math, multilingual and multi-turn chat tasks; it significantly outperforms existing models smaller or similar in size.\n\nThis model is ready for commercial use. \n\n### Key features\n1. Released under the Apache 2 License\n2. Pre-trained and instructed versions\n3. Trained with a 128k context window\n4. Trained on a large proportion of multilingual and code data\n5. Drop-in replacement of Mistral 7B\n\n## Joint-Party Community Consideration\nThis model was a jointly trained by Mistral and NVIDIA.\n\n## License \u0026 Terms of use\nYour use of this API is governed by [the NVIDIA API Trial Service Terms of Use](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf); and the use of this model is governed by [the NVIDIA AI Foundation Models Community License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/). Mistral NeMo-12B is released under the Apache 2.0 license.\n\n## References(s):\nMistral NeMo 12B [Blogpost](https://mistral.ai/news/mistral-nemo/)\n\n## Model Architecture:\n**Architecture Type:** Transformer \u003cbr\u003e\n**Network Architecture:** Mistral \u003cbr\u003e\n**Model Version:** 0.1 \u003cbr\u003e\n\nThis transformer model has the following characteristics:\n* Layers: 40\n* Dim: 5,120\n* Head dim: 128\n* Hidden dim: 14,436\n* Activation Function: SwiGLU\n* Number of heads: 32\n* Number of kv-heads: 8 (GQA)\n* Rotary embeddings (theta = 1M)\n* Vocabulary size: 2**17 ~= 128k\n\n**Input** \n* Input Type: Text\n* Input Format: String\n* Input Parameters: max_tokens, temperature, top_p, stop, frequency_penalty, presence_penalty, seed\n\n**Output** \n* Output Type: Text\n* Output Format: String\n\n## Software Integration:\n* Supported Hardware Platform(s): NVIDIA Hopper \n* Preferred Operating System(s): Linux \u003cbr\u003e\n\n### Benchmarks\n#### Main benchmarks\n- HellaSwag (0-shot): 83.5%\n- Winogrande (0-shot): 76.8%\n- OpenBookQA (0-shot): 60.6%\n- CommonSenseQA (0-shot): 70.4%\n- TruthfulQA (0-shot): 50.3%\n- MMLU (5-shot): 68.0%\n- TriviaQA (5-shot): 73.8%\n- NaturalQuestions (5-shot): 31.2%\n#### Multilingual benchmarks\n- MMLU\n - French: 62.3%\n - German: 62.7%\n - Spanish: 64.6%\n - Italian: 61.3%\n - Portuguese: 63.3%\n - Russian: 59.2%\n - Chinese: 59.0%\n - Japanese: 59.0%\n#### Instruct benchmarks\n - MT Bench (dev): 7.84\n - MixEval Hard: 0.534\n - IFEval-v5: 0.629\n - Wildbench: 42.57\n\n## Inference\n**Engine:** TensorRT-LLM \u003cbr\u003e\n**Test Hardware:** H100\n\n## Ethical Considerations:\nWhen downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"a0:T8ef,"])</script><script>self.__next_f.push([1,"{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Tell me about Dumbledore.\"\n }\n ],\n \"model\": \"nv-mistralai/mistral-nemo-12b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"describe_harry_potter_character\",\n \"description\": \"Returns information and images of Harry Potter characters.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"name\": {\n \"type\": \"string\",\n \"enum\": [\n \"Harry James Potter\",\n \"Hermione Jean Granger\",\n \"Ron Weasley\",\n \"Fred Weasley\",\n \"George Weasley\",\n \"Bill Weasley\",\n \"Percy Weasley\",\n \"Charlie Weasley\",\n \"Ginny Weasley\",\n \"Molly Weasley\",\n \"Arthur Weasley\",\n \"Neville Longbottom\",\n \"Luna Lovegood\",\n \"Draco Malfoy\",\n \"Albus Percival Wulfric Brian Dumbledore\",\n \"Minerva McGonagall\",\n \"Remus Lupin\",\n \"Rubeus Hagrid\",\n \"Sirius Black\",\n \"Severus Snape\",\n \"Bellatrix Lestrange\",\n \"Lord Voldemort\",\n \"Cedric Diggory\",\n \"Nymphadora Tonks\",\n \"James Potter\"\n ],\n \"description\": \"Name of the Harry Potter character\"\n }\n },\n \"required\": [\n \"name\"\n ]\n }\n }\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"a1:T541,{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"What is the weather in Santa Clara, CA?\"\n }\n ],\n \"model\": \"nv-mistralai/mistral-nemo-12b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_current_weather\",\n \"description\": \"A tool that gets the current weather at a location, if one is specified, and defaults to the user's location.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"location\": {\n \"type\": \"string\",\n \"description\": \"The location to find the weather of, or if not provided, it's the default location.\"\n },\n \"unit\": {\n \"type\": \"string\",\n \"enum\": [\n \"u\",\n \"m\"\n ],\n \"description\": \"Whether to use SI or USCS units (celsius or fahrenheit). Infer this from the user's location.\"\n }\n }\n }\n }\n }\n ]\n}\na2:T4a6,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nv-mistralai/mistral-nemo-12b-instruct:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API cal"])</script><script>self.__next_f.push([1,"l using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"mistral-nemo-12b-instruct\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).a3:Tfdc,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/support-matrix.html#llama-3-2-nv-embedqa-1b-v2))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Experience via App\n\n| [ChatRTX](https://www.nvidia.com/en-us/ai-on-rtx/chatrtx/) | [AnythingLLM](https://docs.anythingllm.com/nvidia-nims/introduction) | [AI Toolkit for VSCode](https://aka.ms/aitoolkit)|\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench -u root\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod -R a+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nv-mistralai/mistral-nemo-12b-instruct:1.8.0-rtx\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"nvidia-mistralai/mistral-nemo-12b-instruct\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html)\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\n"])</script><script>self.__next_f.push([1,"a4:T5777,"])</script><script>self.__next_f.push([1,"## Model Information\n\nThe Meta Llama 3.1 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction tuned generative models in 8B, 70B and 405B sizes (text in/text out). The Llama 3.1 instruction tuned text only models (8B, 70B, 405B) are optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks.\n\n**Model Developer**: Meta\n\n## Llama 3.1 Systems\n\n**Large language models, including Llama 3.1, are not designed to be deployed in isolation but instead should be deployed as part of an overall AI system with additional safety guardrails as required.** Developers are expected to deploy system safeguards when building agentic systems. Safeguards are key to achieve the right helpfulness-safety alignment as well as mitigating safety and security risks inherent to the system and any integration of the model or system with external tools. \nAs part of our responsible release approach, we provide the community with [safeguards](https://llama.meta.com/trust-and-safety/) that developers should deploy with Llama models or other LLMs, including Llama Guard 3, Prompt Guard and Code Shield. All our [reference implementations](https://github.com/meta-llama/llama-agentic-system) demos contain these safeguards by default so developers can benefit from system-level safety out-of-the-box.\n\n## Intended Use\n\n**Intended Use Cases** Llama 3.1 is intended for commercial and research use in multiple languages. Instruction tuned text only models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks. The Llama 3.1 model collection also supports the ability to leverage the outputs of its models to improve other models including synthetic data generation and distillation. The Llama 3.1 Community License allows for these use cases.\n\n**Out-of-scope** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in any other way that is prohibited by the Acceptable Use Policy and Llama 3.1 Community License. Use in languages beyond those explicitly referenced as supported in this model card**. \n\n**Note: Llama 3.1 has been trained on a broader collection of languages than the 10 supported languages. \n\nDevelopers may fine-tune Llama 3.1 models for languages beyond the 8 supported languages provided they comply with the Llama 3.1 Community License and the Acceptable Use Policy and in such cases are responsible for ensuring that any uses of Llama 3.1 in additional languages is done in a safe and responsible manner.\n\n\n## New Capabilities\n\nNote that this release introduces new capabilities, including a longer context window, multilingual inputs and outputs and possible integrations by developers with third party tools. Building with these new capabilities requires specific considerations in addition to the best practices that generally apply across all Generative AI use cases. \n\n**Tool-use:** Just like in standard software development, developers are responsible for the integration of the LLM with the tools and services of their choice. They should define a clear policy for their use case and assess the integrity of the third party services they use to be aware of the safety and security limitations when using this capability. Refer to the Responsible Use Guide for best practices on the safe deployment of the third party safeguards. \n\n**Multilinguality:** Llama 3.1 supports 7 languages in addition to English: French, German, Hindi, Italian, Portuguese, Spanish, and Thai. Llama may be able to output text in other languages than those that meet performance thresholds for safety and helpfulness. We strongly discourage developers from using this model to converse in non-supported languages without implementing finetuning and system controls in alignment with their policies and the best practices shared in the Responsible Use Guide.\n\n**Model Architecture:** Llama 3.1 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback\n(RLHF) to align with human preferences for helpfulness and safety.\n\n| | Training Data | Params | Input modalities | Output modalities | Context Length | GQA | Token count | Knowledge cutoff |\n|-|-|-----------------------|----------------------------------------------|-----------------------|---------------------|-----------------------|-------|---------------|\n| | | 8B | Multilingual Text | Multilingual Text and code| 128k | Yes | 15T+ | December 2023 |\n| Llama 3.1 (text only) | A new mix of publicly available online data. | 70B | Multilingual Text | Multilingual Text and code| 128k | Yes | 15T+ | December 2023 |\n| | | 405B | Multilingual Text | Multilingual Text and code| 128k | Yes | 15T+ | December 2023 |\n\n**Supported languages:** English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n**Llama 3.1 family of models**. Token counts refer to pretraining data only. All model versions use Grouped-Query Attention (GQA) for improved inference scalability.\n\n**Model Release Date:** July 23, 2024. \n\n**Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback. \n\n**License** A custom commercial license, the Llama 3.1 Community License, is available at: https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/LICENSE \n\nWhere to send questions or comments about the model Instructions on how to provide feedback or comments on the model can be found in the model [README](https://github.com/meta-llama/llama3). For more technical information about generation parameters and recipes for how to use Llama 3.1 in applications, please go [here](ttps://github.com/meta-llama/llama-recipes).\n\n## Hardware And Software\n\n**Training Factors** We used custom training libraries, Meta's custom built GPU cluster, and production infrastructure for pretraining. Fine-tuning, annotation, and evaluation were also performed on production infrastructure. \n\n**Training Energy Use** Training utilized a cumulative of **39.3**M GPU hours of computation on H100-80GB (TDP of 700W) type hardware, per the table below. Training time is the total GPU time required for training each model and power consumption is the peak power capacity per GPU device used, adjusted for power usage efficiency.\n\n**Training Greenhouse Gas Emissions** Estimated total location-based greenhouse gas emissions were **11,390** tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy, therefore the total market-based greenhouse gas emissions for training were 0 tons CO2eq.\n\n| | Training Time (GPU hours) | Training Power Consumption (W) | Training Location-Based Greenhouse Gas Emissions (tons CO2eq) | Training Market-Based Greenhouse Gas Emissions (tons CO2eq) |\n| - |---------------------------------------|---------------------------------------|---------------------------|--------|\n| Llama 3.1 8B | 1.46M | 700 | 420 | 0 |\n| Llama 3.1 70B | 7.0M | 700 | 2,040 | 0 |\n| Llama 3.1 405B | 30.84M | 700 | 8,930 | 0 |\n| Total | 39.3M | - | 11,390 | 0 |\n\nThe methodology used to determine training energy use and greenhouse gas emissions can be found [here](https://arxiv.org/pdf/2204.05149). Since Meta is openly releasing these models, the training energy use and greenhouse gas emissions will not be incurred by others.\n\n## Training Data\n\n**Overview:** Llama 3.1 was pretrained on ~15 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over 25M synthetically generated examples.\n\n**Data Freshness:** The pretraining data has a cutoff of December 2023.\n\n## Benchmarks - English Text\n\nIn this section, we report the results for Llama 3.1 models on standard automatic benchmarks. For all the evaluations, we use our internal evaluations library.\n\n### Base pretrained models\n| Category | Benchmark | # Shots | Metric | Llama 3 8B | Llama 3.1 8B | Llama 3 70B | Llama 3.1 70B | Llama 3.1 405B |\n|--------------------------|---------------|--------------------|----------|------------|--------------|-------------|---------------|----------------|\n| General | MMLU | 5 | macro_avg/acc_char | 66.7 | 66.7 | 79.5 | 79.3 | 85.2 | |\n| General | MMLU PRO (CoT) | 5 | macro_avg/acc_char | 36.2 | 37.1 | 55.0 | 53.8 | 61.6 | |\n| General | AGIEval English | 3-5 | average/acc_char | 47.1 | 47.8 | 63.0 | 64.6 | 71.6 | |\n| General | CommonSenseQA | 7 | acc_char | 72.6 | 75.0 | 83.8 | 84.1 | 85.8 |\n| General | Winogrande | 5 | acc_char | - | 60.5 | - | 83.3 | 86.7 | |\n| General | BIG-Bench Hard (CoT) | 3 | average/em | 61.1 | 64.2 | 81.3 | 81.6 | **85.9** | |\n| General | ARC-Challenge | 25 | acc_char | 79.4 | 79.7 | 93.1 | 92.9 | 96.1 | |\n| Knowledge reasoning | TriviaQA-Wiki | 5 | em | 78.5 | 77.6 | 89.7 | 89.8 | 91.8 |\n| Reading comprehension | SQuAD | 1 | em | 76.4 | 77.0 | 85.6 | 81.8 | 89.3 | |\n| Reading comprehension | QuAC (F1) | 1 | f1 | 44.4 | 44.9 | 51.1 | 51.1 | 53.6 | |\n| Reading comprehension | BoolQ | 0 | acc_char | 75.7 | 75.0 | 79.0 | 79.4 | 80.0 |\n| Reading comprehension | DROP (F1) | 3 | f1 | 58.4 | 59.5 | 79.7 | 79.6 | **84.8** | |\n\n### Instruction Tuned Models\n\n\n| Category | Benchmark | # Shots | Metric | Llama 3 8B Instruct | Llama 3.1 8B Instruct | Llama 3 70B Instruct | Llama 3.1 70B Instruct | Llama 3.1 405B Instruct | \n| --- | --- | --- | --- | --- | --- | --- | --- | --- | \n | General | MMLU | 5 | macro_avg/acc | 68.5 | 69.4 | 82.0 | 83.6 | 87.3 | \n | General | MMLU (CoT) | 0 | macro_avg/acc | 65.3 | 72.7 | 80.9 | 85.9 | 88.6 | \n | General | MMLU PRO (CoT) | 5 | micro_avg/acc_char | 45.5 | 48.3 | 63.4 | 65.1 | 73.3 | \n | Reasoning | ARC-C | 0 | acc | 82.4 | 83.4 | 94.4 | 94.8 | **96.9** | \n | Reasoning | GPQA | 0 | em | 34.6 | 30.4 | 39.5 | 41.7 | 50.7 | \n | Reasoning | MuSR | 0 | correct | 56.3 | 45.7 | 55.1 | 58.1 | 56.7 | \n | Steerability | IFEval | | | 76.8 | 80.4 | 82.9 | 87.5 | **88.6** | \n | Code | HumanEval | 0 | pass@1 | 60.4 | 72.6 | 81.7 | 80.5 | 89.0 | \n | Code | MBPP ++ base version | 0 | pass@1 | 70.6 | 72.8 | 82.5 | 86.0 | 88.6 | \n | Math | GSM-8K (CoT) | 8 | em_maj1@1 | 80.6 | 84.5 | 93.0 | 95.1 | 96.8 | \n | Math | MATH (CoT) | 0 | final_em | 29.1 | 51.9 | 51.0 | 68.0 | 73.8 | \n | Tool Use | API-Bank | 0 | acc | 83.6 | 82.6 | 85.1 | 90.0 | 92.0 | \n | Tool Use | Berkeley Function Calling | 0 | acc | 76.1 | 76.1 | 83.0 | 85.1 | **88.5** |\n | Tool Use | Gorilla Benchmark API Bench | 0 | acc | 8.8 | 8.2 | 14.7 | 29.7 | 35.3 | \n | Tool Use | Nexus (0-shot) | 0 | macro_avg/acc | 37.6 | 38.5 | 47.8 | 56.7 | **58.7** | \n | Multilingual | Multilingual MGSM | 8 | em | - | 68.2 | - | 85.6 | 90.3 |\n\n## Multilingual Benchmarks\n\n| Category | Benchmark | Language | Llama 3.1 8B | Llama 3.1 70B | Llama 3.1 405B | \n| --- | --- | --- | --- | --- | --- | \n| | | Portuguese | 62.12 | 80.13 | 84.95 |\n| | | Spanish | 62.45 | 80.05 | 85.08 |\n| | | Italian | 61.63 | 80.4 | 85.04 | \n| General | MMLU (5-shot, macro_avg/acc) | German | 60.59 | 79.27 | 84.36 | \n| | | French | 62.34 | 79.82 | 84.66 | \n| | | Hindi | 50.88 | 74.52 | 80.31 | \n| | | Thai | 50.32 | 72.95 | 78.21 |\n\n\n\n## Responsibility \u0026 Safety\n\nAs part of our Responsible release approach, we followed a three-pronged strategy to managing trust \u0026 safety risks:\n- Enable developers to deploy helpful, safe and flexible experiences for their target audience and for the use cases supported by Llama.\n\n- Protect developers against adversarial users aiming to exploit Llama capabilities to potentially cause harm.\n\n- Provide protections for the community to help prevent the misuse of our models.\n\n## Responsible Deployment\n\nLlama is a foundational technology designed to be used in a variety of use cases, examples on how Meta's Llama models have been responsibly deployed can be found in our [Community Stories webpage](https://llama.meta.com/community-stories/). Our approach is to build the most helpful models enabling the world to benefit from the technology power, by aligning our model safety for the generic use cases addressing a standard set of harms. Developers are then in the driver seat to tailor safety for their use case, defining their own policy and deploying the models with the necessary safeguards in their Llama systems. Llama 3.1 was developed following the best practices outlined in our Responsible Use Guide, you can refer to the [Responsible Use Guide](https://llama.meta.com/responsible-use-guide/) to learn more.\n\n## Llama 3.1 Instruct\n\nOur main objectives for conducting safety fine-tuning are to provide the research community with a valuable resource for studying the robustness of safety fine-tuning, as well as to offer developers a readily available, safe, and powerful model for various applications to reduce the developer workload to deploy safe AI systems. For more details on the safety mitigations implemented please read the Llama 3 paper.\n\n### Fine-Tuning Data\n\nWe employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks. We've developed many large language model (LLM)-based classifiers that enable us to thoughtfully select high-quality prompts and responses, enhancing data quality control.\n\n### Refusals And Tone\n\nBuilding on the work we started with Llama 3, we put a great emphasis on model refusals to benign prompts as well as refusal tone. We included both borderline and adversarial prompts in our safety data strategy, and modified our safety data responses to follow tone guidelines.\n\n## Evaluations\n\nWe evaluated Llama models for common use cases as well as specific capabilities. Common use cases evaluations measure safety risks of systems for most commonly built applications including chat bot, coding assistant, tool calls. We built dedicated, adversarial evaluation datasets and evaluated systems composed of Llama models and Llama Guard 3 to filter input prompt and output response. It is important to evaluate applications in context, and we recommend building dedicated evaluation dataset for your use case. Prompt Guard and Code Shield are also available if relevant to the application. \n\nCapability evaluations measure vulnerabilities of Llama models inherent to specific capabilities, for which were crafted dedicated benchmarks including long context, multilingual, tools calls, coding or memorization.\n\n## Red Teaming\n\nFor both scenarios, we conducted recurring red teaming exercises with the goal of discovering risks via adversarial prompting and we used the learnings to improve our benchmarks and safety tuning datasets. We partnered early with subject-matter experts in critical risk areas to understand the nature of these real-world harms and how such models may lead to unintended harm for society. Based on these conversations, we derived a set of adversarial goals for the red team to attempt to achieve, such as extracting harmful information or reprogramming the model to act in a potentially harmful capacity. The red team consisted of experts in cybersecurity, adversarial machine learning, responsible AI, and integrity in addition to multilingual content specialists with background in integrity issues in specific geographic markets. .\n\n## Critical And Other Risks\n\nWe specifically focused our efforts on mitigating the following critical risk areas: \n\n ### 1- CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosive materials) helpfulness\n To assess risks related to proliferation of chemical and biological weapons, we performed uplift testing designed to assess whether use of Llama 3.1 models could meaningfully increase the capabilities of malicious actors to plan or carry out attacks using these types of weapons.\n\n### 2. Child Safety\n\nChild Safety risk assessments were conducted using a team of experts, to assess the model's capability to produce outputs that could result in Child Safety risks and inform on any necessary and appropriate risk mitigations via fine tuning. We leveraged those expert red teaming sessions to expand the coverage of our evaluation benchmarks through Llama 3 model development. For Llama 3, we conducted new in-depth sessions using objective based methodologies to assess the model risks along multiple attack vectors including the additional languages Llama 3 is trained on. We also partnered with content specialists to perform red teaming exercises assessing potentially violating content while taking account of market specific nuances or experiences.\n\n### 3. Cyber Attack Enablement\n\nOur cyber attack uplift study investigated whether LLMs can enhance human capabilities in hacking tasks, both in terms of skill level and speed. Our attack automation study focused on evaluating the capabilities of LLMs when used as autonomous agents in cyber offensive operations, specifically in the context of ransomware attacks. This evaluation was distinct from previous studies that considered LLMs as interactive assistants. The primary objective was to assess whether these models could effectively function as independent agents in executing complex cyber-attacks without human intervention. Our study of Llama-3.1-405B's social engineering uplift for cyber attackers was conducted to assess the effectiveness of AI models in aiding cyber threat actors in spear phishing campaigns. Please read our Llama 3.1 Cyber security whitepaper to learn more.\n\n## Community\n\nGenerative AI safety requires expertise and tooling, and we believe in the strength of the open community to accelerate its progress. We are active members of open consortiums, including the AI Alliance, Partnership on AI and MLCommons, actively contributing to safety standardization and transparency. We encourage the community to adopt taxonomies like the MLCommons Proof of Concept evaluation to facilitate collaboration and transparency on safety and content evaluations. Our Purple Llama tools are open sourced for the community to use and widely distributed across ecosystem partners including cloud service providers. We encourage community contributions to our [Github repository](https://github.com/meta-llama/PurpleLlama). \n\nWe also set up the [Llama Impact Grants](https://llama.meta.com/llama-impact-grants/) program to identify and support the most compelling applications of Meta's Llama model for societal benefit across three categories: education, climate and open innovation. The 20 finalists from the hundreds of applications can be found [here](https://llama.meta.com/llama-impact-grants/#finalists). \nFinally, we put in place a set of resources including an [output reporting mechanism](https://developers.facebook.com/llama_output_feedback) and [bug bounty program](https://www.facebook.com/whitehat) to continuously improve the Llama technology with the help of the community.\n\n## Ethical Considerations And Limitations\n\nThe core values of Llama 3.1 are openness, inclusivity and helpfulness. It is meant to serve everyone, and to work for a wide range of use cases. It is thus designed to be accessible to people across many different backgrounds, experiences and perspectives. Llama 3.1 addresses users and their needs as they are, without insertion unnecessary judgment or normativity, while reflecting the understanding that even content that may appear problematic in some cases can serve valuable purposes in others. It respects the dignity and autonomy of all users, especially in terms of the values of free thought and expression that power innovation and progress. \n\nBut Llama 3.1 is a new technology, and like any new technology, there are risks associated with its use. Testing conducted to date has not covered, nor could it cover, all scenarios. For these reasons, as with all LLMs, Llama 3.1's potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 3.1 models, developers should perform safety testing and tuning tailored to their specific applications of the model. Please refer to available resources including our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide), [Trust and Safety](https://llama.meta.com/trust-and-safety/) solutions, and other [resources](https://llama.meta.com/docs/get-started/) to learn more about responsible development."])</script><script>self.__next_f.push([1,"a5:T8e3,"])</script><script>self.__next_f.push([1,"{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Tell me about Dumbledore.\"\n }\n ],\n \"model\": \"meta/llama-3.1-8b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"describe_harry_potter_character\",\n \"description\": \"Returns information and images of Harry Potter characters.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"name\": {\n \"type\": \"string\",\n \"enum\": [\n \"Harry James Potter\",\n \"Hermione Jean Granger\",\n \"Ron Weasley\",\n \"Fred Weasley\",\n \"George Weasley\",\n \"Bill Weasley\",\n \"Percy Weasley\",\n \"Charlie Weasley\",\n \"Ginny Weasley\",\n \"Molly Weasley\",\n \"Arthur Weasley\",\n \"Neville Longbottom\",\n \"Luna Lovegood\",\n \"Draco Malfoy\",\n \"Albus Percival Wulfric Brian Dumbledore\",\n \"Minerva McGonagall\",\n \"Remus Lupin\",\n \"Rubeus Hagrid\",\n \"Sirius Black\",\n \"Severus Snape\",\n \"Bellatrix Lestrange\",\n \"Lord Voldemort\",\n \"Cedric Diggory\",\n \"Nymphadora Tonks\",\n \"James Potter\"\n ],\n \"description\": \"Name of the Harry Potter character\"\n }\n },\n \"required\": [\n \"name\"\n ]\n }\n }\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"a6:T535,{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"What is the weather in Santa Clara, CA?\"\n }\n ],\n \"model\": \"meta/llama-3.1-8b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_current_weather\",\n \"description\": \"A tool that gets the current weather at a location, if one is specified, and defaults to the user's location.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"location\": {\n \"type\": \"string\",\n \"description\": \"The location to find the weather of, or if not provided, it's the default location.\"\n },\n \"unit\": {\n \"type\": \"string\",\n \"enum\": [\n \"u\",\n \"m\"\n ],\n \"description\": \"Whether to use SI or USCS units (celsius or fahrenheit). Infer this from the user's location.\"\n }\n }\n }\n }\n }\n ]\n}\na7:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n "])</script><script>self.__next_f.push([1," model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\na8:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();a9:T66d,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta/llam"])</script><script>self.__next_f.push([1,"a-3.1-8b-instruct\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003eaa:T49b,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/meta/llama-3.1-8b-instruct:latest\n```\n\n## Step 3. "])</script><script>self.__next_f.push([1,"Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"meta/llama-3.1-8b-instruct\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).ab:Tf7d,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/large-language-models/latest/supported-models.html))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Experience via App\n\n| [ChatRTX](https://www.nvidia.com/en-us/ai-on-rtx/chatrtx/) | [AnythingLLM](https://docs.anythingllm.com/nvidia-nims/introduction) | [AI Toolkit for VSCode](https://aka.ms/aitoolkit)|\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod -R a+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=8GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/meta/llama-3.1-8b-instruct:1.8.0-RTX\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"meta/llama-3.1-8b-instruct\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Hello! How are you?\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\n"])</script><script>self.__next_f.push([1,"ac:Te38,"])</script><script>self.__next_f.push([1,"# Model Overview\nParakeet-CTC-XL-0.6B (around 600M parameters) is trained on ASRSet with over 35000 hours of English (en-US) speech. The model transcribes speech in lower case English alphabet along with spaces and apostrophes.\n\n## Description\nParakeet transcribes audio into text, using spaces and apostrophes where needed \u003cbr\u003e\n\n## Terms of use\nBy using this software or microservice, you are agreeing to the [terms and conditions](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) of the license and acceptable use policy.\n\n# Disclaimer\nAI models generate responses and outputs based on complex algorithms and machine learning techniques, and those responses or outputs may be inaccurate or indecent. By testing this model, you assume the risk of any harm caused by any response or output of the model. Please do not upload any confidential information or personal data. Your use is logged for security. \n\n# References\n* [Paper](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast-conformer)\n* [Fast-Conformer-CTC Model](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/models.html#fast-conformer)\n* [Conformer](https://docs.nvidia.com/nemo-framework/user-guide/latest/nemotoolkit/asr/models.html#conformer)\n\n## Model Architecture\n**Architecture Type:** Parakeet-CTC (also known as FastConformer-CTC), which is an optimized version of Conformer model with 8x depthwise-separable convolutional downsampling with CTC loss \u003cbr\u003e \n**Network Architecture:** Parakeet-CTC-XL-0.6B \u003cbr\u003e\n\n## Input\n**Input Type(s):** Audio in English \u003cbr\u003e\n**Input Format(s):** Linear PCM 16-bit 1 channel \u003cbr\u003e\n**Other Properties Related to Input:** Maximum Length in seconds specific to GPU Memory, No Pre-Processing Needed, Mono channel is required \u003cbr\u003e\n\n## Output\n**Output Type(s):** Text String in English with Timestamps \u003cbr\u003e\n**Output Parameters:** 1-Dimension \u003cbr\u003e\n**Other Properties Related to Output:** No Maximum Character Length, Does not handle special characters \u003cbr\u003e\n\n\n## Software Integration\n**Runtime Engine(s):** \n* Riva 2.15.0 or Higher \u003cbr\u003e\n\n**Supported Operating System(s):** \u003cbr\u003e\n* Linux \u003cbr\u003e\n\n## Model Version\nParakeet-0.6b-ctc-en-us-asr-set-6.0 \u003cbr\u003e\n\n# Inference\n**Engine:** Triton \u003cbr\u003e\n**Test Hardware:** \u003cbr\u003e\n* NVIDIA A10 \u003cbr\u003e\n* NVIDIA A100 \u003cbr\u003e\n* NVIDIA A30 \u003cbr\u003e\n* NVIDIA H100 \u003cbr\u003e\n* NVIDIA Jetson Orin \u003cbr\u003e\n* NVIDIA L4 \u003cbr\u003e\n* NVIDIA L40 \u003cbr\u003e\n* NVIDIA Turing T4 \u003cbr\u003e\n* NVIDIA Volta V100 \u003cbr\u003e\n\n\n## Ethical Considerations (For NVIDIA Models Only):\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## GOVERNING TERMS: \nThis trial is governed by the NVIDIA API Trial Terms of Service (found at https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). The use of this model is governed by the AI Foundation Models Community License Agreement (found at NVIDIA Agreements | Enterprise Software | NVIDIA AI Foundation Models Community License Agreement)."])</script><script>self.__next_f.push([1,"ad:T996,"])</script><script>self.__next_f.push([1,"Field | Response\n:------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\nIntended Applications \u0026 Domains: | Speech Transcription\nTypes: | Speech Transcription\nIntended Users: | This model is intended for developers and data scientists building interactive call centers, virtual assistants, language learning assistants.\nOutput: | Transcribed text with timestamps and confidence scores\nDescribe how the model works: | Model transcribes audio input into text for the input language\nName the adversely impacted groups this has been tested to deliver comparable outcomes regardless of: | Age, Gender, National Origin\nTechnical Limitations: | Transcripts may not be 100% accurate. Accuracy varies based on the characteristics of input audio (Domain, Use Case, Accent, Noise, Speech Type, Context of speech, etc.)\nVerified to have met prescribed NVIDIA quality standards: | Yes\nPerformance Metrics: | Word Error Rate (WER), Silence Robustness (Characters/mins of silent audio), Latency (in milliseconds), Throughput (Total audio processed per unit of time)\nPotential Known Risks: | Not recommended for word-for-word transcription as accuracy varies based on the characteristics of input audio (domain, use case, accent, noise, speech type, and context of speech)\nLicensing: | [https://docs.nvidia.com/ai-foundation-models-community-license.pdf](https://docs.nvidia.com/ai-foundation-models-community-license.pdf)"])</script><script>self.__next_f.push([1,"ae:T83d,"])</script><script>self.__next_f.push([1,"Field | Response\n:----------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------\nGeneratable or reverse engineerable personally-identifiable information (PII)? | None\nWas consent obtained for any personal data used? | Yes\nProtected class data used to create this model? | Age, Gender, Linguistic Background, National Origin\nHow often is dataset reviewed? | Before Release\nIs a mechanism in place to honor data subject right of access or deletion of personal data? | No\nIf Personal data collected for the development of the model, was it collected directly by NVIDIA? | Personal data not collected by NVIDIA for development of model\nIf Personal data collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable\nIf Personal data collected for the development of this AI model, was it minimized to only what was required? | Yes\nIs there provenance for all datasets used in training? | Yes\nDoes data labeling (annotation, metadata) comply with privacy laws? | Yes \nIs data compliant with data subject requests for data correction or removal, if such a request was made? | The data is compliant where applicable, but is not applicable for all data."])</script><script>self.__next_f.push([1,"af:T688,### Getting Started\n\nRiva uses \u003ca href=\"https://grpc.io/\"\u003egRPC\u003c/a\u003e APIs. Instructions below demonstrate usage of \u003c%- name %\u003e model using Python gRPC client.\n\n### Prerequisites\n\nYou will need a system with Git and Python 3+ installed.\n\n### Install Riva Python Client\n\n```bash\npip install nvidia-riva-client\n```\n\n### Download Python Client\n\nDownload Python client code by cloning \u003ca href=\"https://github.com/nvidia-riva/python-clients\"\u003ePython Client Repository\u003c/a\u003e.\n\n```bash\ngit clone https://github.com/nvidia-riva/python-clients.git\n```\n\n### Run Python Client\n\nOpen a command terminal and execute below command to transcribe audio. Make sure you have a speech file in 16-bit Mono format in WAV/OGG/OPUS container. If you have generated the API key, it will be auto-populated in the command.\n\n```bash\npython python-clients/scripts/asr/transcribe_file.py \\\n --server grpc.nvcf.nvidia.com:443 --use-ssl \\\n --metadata function-id \"\u003c%- nvcfFunctionId %\u003e\" \\\n --metadata \"authorization\" \"Bearer \u003c%- apiKey %\u003e\" \\\n --language-code en-US \\\n --input-file \u003cpath_to_audio_file\u003e\n```\n\n### Support for gRPC clients in other languages\n\nRiva uses \u003ca href=\"https://grpc.io/\"\u003egRPC\u003c/a\u003e APIs. Proto files can be downloaded from \u003ca href=\"https://github.com/nvidia-riva/common/archive/refs/heads/main.zip\"\u003eRiva gRPC Proto files\u003c/a\u003e and compiled to target language using \u003ca href=\"https://grpc.io/docs/protoc-installation/\"\u003eProtoc compiler\u003c/a\u003e. Example Riva clients in C++ and Python languages are provided below.\n\n* \u003ca href=\"https://github.com/nvidia-riva/python-clients\"\u003ePython Client Repository\u003c/a\u003e\n* \u003ca href=\"https://github.com/nvidia-riva/cpp-clients\"\u003eC++ Client Repository\u003c/a\u003e\nb0:T707,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nRefer [Supported Models](https://docs.nvidia.com/nim/riva/asr/latest/getting-started.html#supported-models) for full list of models.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\n\ndocker"])</script><script>self.__next_f.push([1," run -it --rm \\\n --runtime=nvidia \\\n --gpus '\"device=0\"' \\\n --shm-size=8GB \\\n -e NGC_API_KEY \\\n -e NIM_HTTP_API_PORT=9000 \\\n -e NIM_GRPC_API_PORT=50051 \\\n -p 9000:9000 \\\n -p 50051:50051 \\\n -e NIM_TAGS_SELECTOR=name=parakeet-0-6b-ctc-riva-en-us,mode=ofl,bs1 \\\n nvcr.io/nim/nvidia/parakeet-0-6b-ctc-en-us:latest\n```\n\nIt may take a up to 30 minutes depending on your network speed, for the container to be ready and start accepting requests from the time the docker container is started.\n\n## Step 3. Test the NIM\n\nOpen a new terminal and run following command to check if the service is ready to handle inference requests\n\n```bash\ncurl -X 'GET' 'http://localhost:9000/v1/health/ready'\n```\n\nIf the service is ready, you get a response similar to the following.\n```bash\n{\"ready\":true}\n```\n\nInstall the Riva Python client package\n\n```bash\nsudo apt-get install python3-pip\npip install nvidia-riva-client\n```\n\nDownload Riva sample clients\n\n```bash\ngit clone https://github.com/nvidia-riva/python-clients.git\n```\n\nRun Speech to Text inference in streaming modes. Riva ASR supports Mono, 16-bit audio in WAV, OPUS and FLAC formats.\n\n```bash\npython3 python-clients/scripts/asr/transcribe_file_offline.py --server 0.0.0.0:50051 --input-file \u003cpath_to_speech_file\u003e --language-code en-US\n```\n\n\nFor more details on getting started with this NIM, visit the [Riva ASR NIM Docs](https://docs.nvidia.com/nim/riva/asr/latest/overview.html).b1:Tdbf,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/riva/asr/latest/support-matrix.html)) \n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+) \n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Step 1\\. Open the Windows Subsystem for Linux 2 \\- WSL2 \\- Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the `NVIDIA-Workbench` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2\\. Run the Container\n\n::generate-api-key\n\n```\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below.\n\n```\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod -R a+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -e NIM_TAGS_SELECTOR=name=parakeet-0-6b-ctc-riva-en-us,mode=ofl,bs=1 \\\n -e NIM_HTTP_API_PORT=9000 \\\n -e NIM_GRPC_API_PORT=50051 \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 9000:9000 \\\n -p 50051:50051 \\\n nvcr.io/nim/nvidia/parakeet-0-6b-ctc-en-us:latest\n```\n\nIt may take up to 30 minutes depending on your network speed for the container to be ready and start accepting requests from the time the docker container is started.\n\n## Step 3\\. Test the NIM\n\nOpen a new Distro instance and run following command to check if the service is ready to handle inference requests\n\n```\ncurl -X 'GET' 'http://localhost:9000/v1/health/ready'\n```\n\nIf the service is ready, you get a response similar to the following.\n\n```\n{\"ready\":true}\n```\n\nInstall the Riva Python client package\n\n```\nsudo apt-get install python3-pip\npip install nvidia-riva-client\n```\n\nDownload Riva sample clients\n\n```\ngit clone https://github.com/nvidia-riva/python-clients.git\n```\n\nRun Speech to Text inference in streaming modes. Riva ASR supports Mono, 16-bit audio in WAV, OPUS and FLAC formats.\n\n```\npython3 python-clients/scripts/asr/transcribe_file_offline.py --server 0.0.0.0:50051 --input-file \u003cpath_to_speech_file\u003e --language-code en-US\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/riva/asr/latest/overview.html).\n\n \n"])</script><script>self.__next_f.push([1,"b2:T1a78,"])</script><script>self.__next_f.push([1,"# Model Overview\n\n## Description\n\nMaxine Studio Voice enhances the input speech recorded through low quality \nmicrophones in noisy/reverberant environment to studio-recorded quality speech.\n\nStudio Voice is available under NVIDIA Maxine — a developer platform for \ndeploying AI features that enhance audio, video, and creating new experiences \nin real-time audio-video communication. Maxine's state-of-the-art models create \nhigh-quality AI effects using standard microphones and cameras without \nadditional special equipments.\n\nNVIDIA Maxine is exclusively part of NVIDIA AI Enterprise for production \nworkflows — an extensive library of full-stack software, including AI solution \nworkflows, frameworks, pre-trained models, and infrastructure optimization. \n\n\n## Terms of use\n\nThe use of NVIDIA Maxine's Studio Voice is available as a demonstration of the \ninput and output of the Studio Voice generative model. As such the user may \nupload an audio file or select one of the sample inputs and download the \ngenerated audio for evaluation under the terms of the \n[NVIDIA MAXINE EVALUATION LICENSE AGREEMENT](https://developer.download.nvidia.com/maxine/nvidia-maxine-evaluation-license-24oct2023.pdf). \n\n\n## References(s):\n\n* [NVIDIA Maxine](https://developer.nvidia.com/maxine) \n\n\n## Model Architecture\n\n**Architecture Type:** Convolution Neural Networks (CNNs), Transformers,\n Generative Adversarial Networks (GANs) \n**Network Architecture:** Encoder-Decoder \n**Model Version:** 0.2 \n\n\n## Input:\n\n**Input Type(s):** Ordered List (audio samples) \n**Input Format(s):** FP32 (-1.0 to 1.0) \n**Other Properties Related to Input:** Pulse Code Modulation (PCM) audio samples\nwith no encoding or pre-processing; 16kHz or 48kHz sampling rate required. \n\n## Output:\n\n**Output Type(s):** Ordered List (audio samples) \n**Output Format:** FP32 (-1.0 to 1.0) \n**Other Properties Related to Output:** PCM audio samples at input sampling rate\nwith no encoding or post-processing. \n\n## Software Integration\n\n**Supported Hardware Platform(s):** Hopper, Ada, Ampere, Turing, Volta \n**Test Hardware:** A10, L40, T10 \n**Supported Operating System(s):** Linux, Windows \n\n\n# Training \u0026 Evaluation\n\n## Datasets\n\nNVIDIA models are trained on a diverse set of public and proprietary datasets.\nThe Studio Voice model is trained on a dataset that comprises of diverse \nEnglish accents.\n\n**Link:** [DAPS](https://ccrma.stanford.edu/~gautham/Site/daps.html) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nThe DAPS dataset has 15 versions of audio (3 professional versions and\n12 consumer device/real-world environment combinations). Each version consists\nof about 4.5 hours of data (about 14 minutes from each of 20 speakers). \n\n**Link:** [LibriTTS](https://www.openslr.org/60/) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nLibriTTS is a multi-speaker English corpus of approximately 585 hours of read\nEnglish speech, which is resampled at 16kHZ. \n\n**Link:** [VCTK](https://datashare.ed.ac.uk/handle/10283/3443) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nThis CSTR VCTK Corpus includes speech data uttered by 110 English speakers with\nvarious accents. Each speaker reads out about 400 sentences, which were selected\nfrom a newspaper, the rainbow passage and an elicitation paragraph used for the\nspeech accent archive. \n\n**Link:** [HiFi-TTS](https://www.openslr.org/109/) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nA multi-speaker English dataset for training text-to-speech models.\nThe HiFi-TTS dataset contains about 291.6 hours of speech from 10 speakers with\nat least 17 hours per speaker sampled at 44.1 kHz. \n\n**Link:** [Device Recorded VCTK (DR-VCTK)](https://github.com/nii-yamagishilab/downloader-DR-VCTK-complete) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nDevice recorded version of VCTK dataset on common consumer devices \n(laptop, tablet and smartphone) in office environment. This dataset contains \n109 English speakers with different accents. There are around 400 sentences \navailable from each speaker. For this recording, 8 different microphones were \nused. This dataset contains around 250 Gb of re-recorded speech. \n\n**Link:** [Dataset of impulse responses from variable acoustics room Arni at Aalto Acoustic Labs](https://zenodo.org/records/6985104#.Y7_vv3ZBy3C) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nA dataset of impulse responses collected in the variable acoustics laboratory \nArni at Acoustics Lab of Aalto University, Espoo, Finland. IRs of 5342 \nconfigurations of sound absorption in Arni are included in the dataset. Each of \nthem were measured using an omnidirectional sound source and 5 sound receivers. \nFor each configuration, 5 impulse reponses (IRs) were captured. The total number\nof measurements in the dataset is 132 037. \n\n**Link:** [Room Impulse Response and Noise Database](https://www.openslr.org/28/) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nA database of simulated and real room impulse responses, isotropic and\npoint-source noises. The audio files in this data are all in 16KHz sampling rate\nand 16-bit precision. \n\n**Link:** [DNS Challenge 5](https://github.com/microsoft/DNS-Challenge/tree/2db96d5f75257df764a6ef66513b4b97bc707f30) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nCollated dataset of clean speech, noise and impulse response provided by \nMicrosoft for the ICASSP 2023 Deep Noise Suppression Challenge. \n\n**Link:** [AudioSet](https://research.google.com/audioset/download.html) \n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** \nAudioSet consists of an expanding ontology of 632 audio event classes and \na collection of 2,084,320 human-labeled 10-second sound clips drawn from \nYouTube videos. \n\n\n# Inference\n\n**Engine:** [Triton](https://developer.nvidia.com/triton-inference-server) \n**Test Hardware:** A10, L40, T10 \n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established\npolicies and practices to enable development for a wide array of AI applications.\nWhen downloaded or used in accordance with our terms of service, developers\nshould work with their supporting model team to ensure this model meets\nrequirements for the relevant industry and use case and addresses unforeseen\nproduct misuse. For more detailed information on ethical considerations for this\nmodel, please see the Model Card++ [Explainability](explainability), [Bias](bias),\n[Safety \u0026 Security](safety-and-security), and [Privacy](privacy) Subcards. Please\nreport security vulnerabilities or NVIDIA AI Concerns\n[here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"b3:T59f,| Field | Response |\n|-------|----------|\n| Intended Application \u0026 Domain: | Voice enhancement |\n| Model Type: | Speech Enhancement |\n| Intended Users: | Content developers and Broadcasters |\n| Output: | Audio |\n| Describe how the model works: | This model enhances the voice in input audio by removing acoustic noise, room reverberation, and bad frequency characteristics of the recording device while maintaining the linguistic and vocal properties. |\n| Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of: | Age (18+), Gender |\n| Verified to have met prescribed quality standards: | Yes |\n| Target Key Performance Indicator(s) (KPI(s)): | Quality Mean Opinion Score (MOS), Non-Intrusive Speech Quality Assessment (NISQA), Latency, Throughput |\n| Technical Limitations: | The model may not work work well on a variety of demographic and regional representations of English or with very noisy or very low quality inputs. Exclaimatory emotions may not be transferred to the output. |\n| Potential Known Risks: | The model may not preserve some speaker characteristics like pitch/fundamental frequency of the speakers voice and might also impact intelligibility of speech. |\n| Licensing: | [NVIDIA Maxine Evaluation License Agreement](https://developer.download.nvidia.com/maxine/nvidia-maxine-evaluation-license-24oct2023.pdf) |b4:T40d,| Field | Response |\n|-------|----------|\n| Generatable or reverse engineerable personally-identifiable information (PII)? | None |\n| Was consent obtained for any PII used? | Yes |\n| Protected class data used to create this model? | None |\n| How often is dataset reviewed? | Before Release |\n| Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n| If PII collected for the development of the model, was it collected directly by NVIDIA? | No |\n| If PII collected for the development of the model by NVIDIA, do you mai"])</script><script>self.__next_f.push([1,"ntain or have access to disclosures made to data subjects? | No |\n| If PII collected for the development of this AI model, was it minimized to only what was required? | Yes |\n| Is there provenance for all datasets used in training? | Yes |\n| Does data labeling (annotation, metadata) comply with privacy laws? | Yes |\n| Is data compliant with data subject requests for data correction or removal, if such a request was made? | No |b5:T919,"])</script><script>self.__next_f.push([1,"## Getting Started\n\nNVIDIA Maxine Studio Voice NIM uses gRPC APIs for inferencing requests. Following instructions demonstrate the usage of Maxine Studio Voice NIM model using Python client.\n\n### Prerequisites\n\nYou will need a system with `git` and `Python 3.10+` installed.\n\n### Setup NVIDIA Maxine Studio Voice Python client\n\nDownload the Maxine Studio Voice Python client code by cloning the [NVIDIA Maxine NIM Clients Repository](https://github.com/NVIDIA-Maxine/nim-clients):\n\n```bash\ngit clone https://github.com/NVIDIA-Maxine/nim-clients.git\ncd nim-clients/studio-voice\n```\n\nInstall the dependencies for the NVIDIA Maxine Studio Voice Python client:\n\n```bash\nsudo apt-get install python3-pip\npip install -r requirements.txt\n```\n\n### Run Python Client\n\nNavigate to the scripts directory.\n\n```bash\ncd scripts\n```\n\nSend the gRPC requests\n\n```bash\npython studio_voice.py --preview-mode \\\n --ssl-mode TLS \\\n --target grpc.nvcf.nvidia.com:443 \\\n --function-id 7cf12edb-2181-4947-8b19-2b1c18270588 \\\n --api-key \u003c%- apiKey %\u003e \\\n --input \u003cinput_file_path\u003e \\\n --output \u003coutput_file_path\u003e\n```\n\nNote the requirements for input file:\n\n- The supported format is 16-bit mono channel wav file.\n- The size limit for input file is 32 MB.\n- The duration limit for input file is 6 min.\n\nCommand line arguments:\n\n- `--preview-mode` - Flag to send request to preview NVCF server on https://build.nvidia.com/nvidia/studiovoice/api.\n- `--ssl-mode` - Set the SSL mode to TLS or MTLS. Defaults to no SSL. When running preview, TLS mode must be used with default root certificate.\n- `--target \u003cip:port\u003e` - URI of NIM's gRPC service. Use grpc.nvcf.nvidia.com:443 when hosted on NVCF. (Default: `127.0.0.1:8001`)\n- `--api-key \u003c%- apiKey %\u003e` - NGC API key required for authentication. Utilized when using `TRY API` ignored otherwise.\n- `--function-id \u003cfunction_id\u003e` - Function ID for the feature.\n- `--input \u003cinput_file_path\u003e` - The path to the input audio file. (Default: `../assets/studio_voice_48k_input.wav`)\n- `--output \u003coutput_file_path\u003e` - The path to the output audio file. (Default: `./studio_voice_48k_output.wav`)\n- `--streaming` - Flag to enable grpc streaming mode.\n\nRefer the [Maxine Studio Voice NIM documentation](https://docs.nvidia.com/nim/maxine/studio-voice/latest/index.html) for more information.\n"])</script><script>self.__next_f.push([1,"b6:T1588,"])</script><script>self.__next_f.push([1,"## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\nNVIDIA Maxine Studio Voice NIM uses gRPC APIs for inferencing requests.\n\nA NGC API KEY is required to download the appropriate models and resources when starting the NIM.\n\nIf you are not familiar with how to create the `NGC_API_KEY` environment variable, the simplest way is to export it in your terminal:\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\n```\n\nRun one of the following commands to make the key available at startup:\n\n```bash\n# If using bash\necho \"export NGC_API_KEY=\u003cvalue\u003e\" \u003e\u003e ~/.bashrc\n\n# If using zsh\necho \"export NGC_API_KEY=\u003cvalue\u003e\" \u003e\u003e ~/.zshrc\n```\n\nOther, more secure options include saving the value in a file, so that you can retrieve with `cat $NGC_API_KEY_FILE`, or using a [password manager](https://www.passwordstore.org/).\n\nThe following command launches the Maxine Studio Voice NIM container with the gRPC service. Find reference to runtime parameters for the container [here](https://docs.nvidia.com/nim/maxine/studio-voice/latest/getting-started.html#runtime-parameters-for-the-container).\n\n```bash\ndocker run -it --rm --name=maxine-studio-voice \\\n --runtime=nvidia \\\n --gpus all \\\n --shm-size=8GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -e NIM_MODEL_PROFILE=\u003cnim_model_profile\u003e \\\n -e FILE_SIZE_LIMIT=36700160 \\\n -p 8000:8000 \\\n -p 8001:8001 \\\n nvcr.io/nim/nvidia/maxine-studio-voice:latest\n```\nEnsure you use the appropriate `NIM_MODEL_PROFILE` for your GPU. For more information about `NIM_MODEL_PROFILE`, refer to the the [NIM Model Profile Table](https://docs.nvidia.com/nim/maxine/studio-voice/latest/model-profile-table.html).\n\nPlease note, the flag --gpus all is used to assign all available GPUs to the docker container. This fails on multiple GPU unless all GPUs are same. To assign specific GPU to the docker container (in case of different multiple GPUs available in your machine) use --gpus '\"device=0,1,2...\"'\n\nIf the command runs successfully, you will get an output ending similar to the following:\n\n```bash\nI1126 09:22:21.048202 31 grpc_server.cc:2558] \"Started GRPCInferenceService at 127.0.0.1:9001\"\nI1126 09:22:21.048377 31 http_server.cc:4704] \"Started HTTPService at 127.0.0.1:9000\"\nI1126 09:22:21.089295 31 http_server.cc:362] \"Started Metrics Service at 127.0.0.1:9002\"\nMaxine GRPC Service: Listening to 0.0.0.0:8001\n```\n\nBy default Maxine Studio Voice gRPC service is hosted on port `8001`. You will have to use this port for inferencing requests.\n\n## Step 3. Test the NIM\n\nWe have provided a sample client script file in our GitHub repo.\nThe script could be used to invoke the Docker container using the following instructions.\n\nDownload the Maxine Studio Voice Python client code by cloning the [NVIDIA Maxine NIM Clients Repository](https://github.com/NVIDIA-Maxine/nim-clients):\n\n```bash\ngit clone https://github.com/NVIDIA-Maxine/nim-clients.git\ncd nim-clients/studio-voice\n```\n\nInstall the dependencies for the NVIDIA Maxine Studio Voice Python client:\n\n```bash\nsudo apt-get install python3-pip\npip install -r requirements.txt\n```\nGo to scripts directory\n```bash\ncd scripts\n```\n\nRun the command to send gRPC request (By Default transactional mode)\n```bash\npython studio_voice.py --target \u003ctarget_ip:port\u003e --input \u003cinput_file_path\u003e --output \u003coutput_file_path\u003e\n```\n\nFor streaming mode:\n```bash\npython studio_voice.py --target \u003ctarget_ip:port\u003e --input \u003cinput_file_path\u003e --output \u003coutput_file_path\u003e --streaming --model-type 48k-hq\n```\n\nWhen using `--streaming` mode, ensure the selected `--model-type` (48k-hq, 48k-ll, or 16k-hq) aligns with the `NIM_MODEL_PROFILE` Model Type configuration to maintain compatibility .\n\nTo view details of command line arguments run this command\n```bash\npython studio_voice.py -h\n```\nYou will get a response similar to the following.\n```bash\nusage: studio_voice.py [-h] [--ssl-mode {MTLS,TLS}] [--ssl-key SSL_KEY] [--ssl-cert SSL_CERT] [--ssl-root-cert SSL_ROOT_CERT] [--target TARGET]\n [--input INPUT] [--output OUTPUT] [--api-key API_KEY] [--function-id FUNCTION_ID] [--streaming] [--model-type {48k-hq,48k-ll,16k-hq}]\n\nProcess wav audio files using gRPC and apply studio-voice.\n\noptions:\n -h, --help show this help message and exit\n --preview-mode Flag to send request to preview NVCF server on https://build.nvidia.com/nvidia/studiovoice/api.\n --ssl-mode {MTLS,TLS}\n Flag to set SSL mode, default is None\n --ssl-key SSL_KEY The path to ssl private key.\n --ssl-cert SSL_CERT The path to ssl certificate chain.\n --ssl-root-cert SSL_ROOT_CERT\n The path to ssl root certificate.\n --target TARGET IP:port of gRPC service, when hosted locally. Use grpc.nvcf.nvidia.com:443 when hosted on NVCF.\n --input INPUT The path to the input audio file.\n --output OUTPUT The path for the output audio file.\n --api-key API_KEY NGC API key required for authentication, utilized when using TRY API ignored otherwise\n --function-id FUNCTION_ID\n NVCF function ID for the service, utilized when using TRY API ignored otherwise\n --streaming Flag to enable grpc streaming mode.\n --model-type {48k-hq,48k-ll,16k-hq}\n Studio Voice model type, default is 48k-hq.\n```\n\nFor more details on getting started with this NIM including configuring using parameters, visit the [NVIDIA Maxine Studio Voice NIM Docs](https://docs.nvidia.com/nim/maxine/studio-voice/latest/index.html).\n\n"])</script><script>self.__next_f.push([1,"b7:T1619,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\nWSL2 is required for hosting any NIM. Refer to the official [NVIDIA NIM on WSL2](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html) documentation for setup instructions.\n\nTo run the NIM refer to the docs on [Studio Voice NIM on WSL2](https://docs.nvidia.com/nim/maxine/studio-voice/latest/running-on-wsl.html#running-studio-voice-nim-on-wsl2)\n\n\n## Step 1. Open the Windows Subsystem for Linux 2 WSL2 Distro\n\n[Install](https://docs.nvidia.com/nim/wsl2/latest/) the WSL2 Distro. \n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal. \n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NIM with the command below.\n```bash\npodman run -it --rm --name=studio-voice \\\n --device nvidia.com/gpu=all \\\n --shm-size=8GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -e NIM_MODEL_PROFILE=\u003cnim_model_profile\u003e \\\n -e FILE_SIZE_LIMIT=36700160 \\\n -e STREAMING=false \\\n -p 8000:8000 \\\n -p 8001:8001 \\\n nvcr.io/nim/nvidia/maxine-studio-voice:latest\n```\n\nThe above command is to run the NIM in transactional mode.\nTo run the NIM in streaming mode use `-e STREAMING=true`.\nEnsure you use the appropriate `NIM_MODEL_PROFILE` for your GPU. For more information about `NIM_MODEL_PROFILE`, refer to the the [NIM Model Profile Table](https://docs.nvidia.com/nim/maxine/studio-voice/latest/model-profile-table.html).\n\nPlease note, the flag --gpus all is used to assign all available GPUs to the docker container. This fails on multiple GPU unless all GPUs are same. To assign specific GPU to the docker container (in case of different multiple GPUs available in your machine) use --gpus '\"device=0,1,2...\"'\n\nIf the command runs successfully, you will get an output ending similar to the following:\n\n```bash\nI1126 09:22:21.048202 31 grpc_server.cc:2558] \"Started GRPCInferenceService at 127.0.0.1:9001\"\nI1126 09:22:21.048377 31 http_server.cc:4704] \"Started HTTPService at 127.0.0.1:9000\"\nI1126 09:22:21.089295 31 http_server.cc:362] \"Started Metrics Service at 127.0.0.1:9002\"\nMaxine GRPC Service: Listening to 0.0.0.0:8001\n```\n\nBy default Maxine Studio Voice gRPC service is hosted on port `8001`. You will have to use this port for inferencing requests.\n\n## Step 3. Test the NIM\n\nWe have provided a sample client script file in our GitHub repo.\nThe script could be used to invoke the Docker container using the following instructions.\n\nDownload the Maxine Studio Voice Python client code by cloning the [NVIDIA Maxine NIM Clients Repository](https://github.com/NVIDIA-Maxine/nim-clients):\n\n```bash\ngit clone https://github.com/NVIDIA-Maxine/nim-clients.git\ncd nim-clients/studio-voice\n```\n\nInstall the dependencies for the NVIDIA Maxine Studio Voice Python client:\n\n```bash\nsudo apt-get install python3-pip\npip install -r requirements.txt\n```\nGo to scripts directory\n```bash\ncd scripts\n```\n\nAssuming the client is on the same machine as the NIM, run the following command to send a gRPC request to the NIM.\n\nBy default transactional mode:\n```bash\npython studio_voice.py --target localhost:8001 --input \u003cinput_file_path\u003e --output \u003coutput_file_path\u003e\n```\n\nFor streaming mode:\n```bash\npython studio_voice.py --target localhost:8001 --input \u003cinput_file_path\u003e --output \u003coutput_file_path\u003e --streaming --model-type 48k-hq\n```\nWhen using `--streaming` mode, ensure the selected `--model-type` (48k-hq, 48k-ll, or 16k-hq) aligns with the `NIM_MODEL_PROFILE` Model Type configuration to maintain compatibility .\n\nFor more advance usage of the client, refer to this [documentation](https://docs.nvidia.com/nim/maxine/studio-voice/latest/running-on-wsl.html#running-nim-when-the-client-and-wsl2-host-are-different)\n\nTo view details of command line arguments run this command\n```bash\npython studio_voice.py -h\n```\nYou will get a response similar to the following.\n```bash\nusage: studio_voice.py [-h] [--ssl-mode {MTLS,TLS}] [--ssl-key SSL_KEY] [--ssl-cert SSL_CERT] [--ssl-root-cert SSL_ROOT_CERT] [--target TARGET]\n [--input INPUT] [--output OUTPUT] [--api-key API_KEY] [--function-id FUNCTION_ID] [--streaming] [--model-type {48k-hq,48k-ll,16k-hq}]\n\nProcess wav audio files using gRPC and apply studio-voice.\n\noptions:\n -h, --help show this help message and exit\n --preview-mode Flag to send request to preview NVCF server on https://build.nvidia.com/nvidia/studiovoice/api.\n --ssl-mode {MTLS,TLS}\n Flag to set SSL mode, default is None\n --ssl-key SSL_KEY The path to ssl private key.\n --ssl-cert SSL_CERT The path to ssl certificate chain.\n --ssl-root-cert SSL_ROOT_CERT\n The path to ssl root certificate.\n --target TARGET IP:port of gRPC service, when hosted locally. Use grpc.nvcf.nvidia.com:443 when hosted on NVCF.\n --input INPUT The path to the input audio file.\n --output OUTPUT The path for the output audio file.\n --api-key API_KEY NGC API key required for authentication, utilized when using TRY API ignored otherwise\n --function-id FUNCTION_ID\n NVCF function ID for the service, utilized when using TRY API ignored otherwise\n --streaming Flag to enable grpc streaming mode.\n --model-type {48k-hq,48k-ll,16k-hq}\n Studio Voice model type, default is 48k-hq.\n```\n\nFor more details on getting started with this NIM including configuring using parameters, visit the [NVIDIA Maxine Studio Voice NIM Docs](https://docs.nvidia.com/nim/maxine/studio-voice/latest/index.html)."])</script><script>self.__next_f.push([1,"b8:T25f5,"])</script><script>self.__next_f.push([1,"## **Model Overview**\n\n### **Description**\n\nThe NVIDIA NeMo Retriever Llama3.2 embedding model is optimized for **multilingual and cross-lingual** text question-answering retrieval with **support for long documents (up to 8192 tokens) and dynamic embedding size (Matryoshka Embeddings)**. This model was evaluated on 26 languages: English, Arabic, Bengali, Chinese, Czech, Danish, Dutch, Finnish, French, German, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Norwegian, Persian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, and Turkish.\n\nIn addition to enabling multilingual and cross-lingual question-answering retrieval, this model reduces the data storage footprint by 35x through dynamic embedding sizing and support for longer token length, making it feasible to handle large-scale datasets efficiently.\n\nAn embedding model is a crucial component of a text retrieval system, as it transforms textual information into dense vector representations. They are typically transformer encoders that process tokens of input text (for example: question, passage) to output an embedding.\n\nThis model is ready for commercial use.\n\nThe Llama 3.2 1b embedding model is a part of the NVIDIA NeMo Retriever collection of NIM, which provide state-of-the-art, commercially-ready models and microservices, optimized for the lowest latency and highest throughput. It features a production-ready information retrieval pipeline with enterprise support. The models that form the core of this solution have been trained using responsibly selected, auditable data sources. With multiple pre-trained models available as starting points, developers can also readily customize them for domain-specific use cases, such as information technology, human resource help assistants, and research \u0026 development research assistants.\n\n### **Intended use**\n\nThe NeMo Retriever Llama3.2 embedding model is most suitable for users who want to build a multilingual question-and-answer application over a large text corpus, leveraging the latest dense retrieval technologies.\n\n### **License/Terms of use**\n\nThe use of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/) and Llama 3.2 is licensed under the [Llama 3.2 Community License](https://www.llama.com/llama3_2/license/), Copyright © Meta Platforms, Inc. All Rights Reserved.\n\n**You are responsible for ensuring that your use of NVIDIA AI Foundation Models complies with all applicable laws.**\n\n### **Model Architecture**\n\n**Architecture Type:** Transformer\n**Network Architecture:** Fine-tuned Llama3.2 1b retriever\n\nThis NeMo Retriever embedding model is a transformer encoder - a fine-tuned version of Llama3.2 1b, with 16 layers and an embedding size of 2048, which is trained on public datasets. The AdamW optimizer is employed incorporating 100 warm up steps and 5e-6 learning rate with WarmupDecayLR scheduler. Embedding models for text retrieval are typically trained using a bi-encoder architecture. This involves encoding a pair of sentences (for example, query and chunked passages) independently using the embedding model. Contrastive learning is used to maximize the similarity between the query and the passage that contains the answer, while minimizing the similarity between the query and sampled negative passages not useful to answer the question.\n\n### **Input**\n\n**Input Type:** Text\n**Input Format:** List of strings\n**Input Parameter:** 1D\n**Other Properties Related to Input:** The model's maximum context length is 8192 tokens. Texts longer than maximum length must either be chunked or truncated.\n\n### **Output**\n\n**Output Type:** Floats\n**Output Format:** List of float arrays\n**Output:** Model outputs embedding vectors of maximum dimension 2048 for each text string (can be configured based on 384, 512, 768, 1024, or 2048).\n**Other Properties Related to Output:** N/A\n\n### **Software Integration**\n\n**Runtime Engine:** NeMo Retriever embedding NIM\n**Supported Hardware Microarchitecture Compatibility**: NVIDIA Ampere, NVIDIA Hopper, NVIDIA Lovelace\n**Supported Operating System(s):** Linux\n\n### **Model Version(s)**\n\nNVIDIA NeMo Retriever Llama 3.2 embedding\nShort Name: llama-3.2-nv-embedqa-1b-v2\n\n## **Training Dataset \u0026 Evaluation**\n\n### **Training Dataset**\n\nThe development of large-scale public open-QA datasets has enabled tremendous progress in powerful embedding models. However, one popular dataset named MS MARCO restricts ‌commercial licensing, limiting the use of these models in commercial settings. To address this, NVIDIA created its own training dataset blend based on public QA datasets, which each have a license for commercial applications.\n\n**Data Collection Method by dataset**: Automated, Unknown\n\n\n**Labeling Method by dataset**: Automated, Unknown\n\n\n**Properties:** Semi-supervised pre-training on 12M samples from public datasets and fine-tuning on 1M samples from public datasets.\n\n\n### **Evaluation Results**\n\nProperties: We evaluated the NeMo Rtriever embdding model in comparison to literature open \u0026 commercial retriever models on academic benchmarks for question-answering - [NQ](https://huggingface.co/datasets/BeIR/nq), [HotpotQA](https://huggingface.co/datasets/hotpot_qa) and [FiQA (Finance Q\\\u0026A)](https://huggingface.co/datasets/BeIR/fiqa) from BeIR benchmark and TechQA dataset. Note that the model was evaluated offline on A100 GPUs using the model's PyTorch checkpoint. In this benchmark, the metric used was Recall@5.\n\n| Open \u0026 Commercial Retrieval Models | Average Recall@5 on NQ, HotpotQA, FiQA, TechQA dataset |\n| ----- | ----- |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 2048) | 68.60% |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 384) | 64.48% |\n| llama-3.2-nv-embedqa-1b-v1 (embedding dim 2048) | 68.97% |\n| nv-embedqa-mistral-7b-v2 | 72.97% |\n| nv-embedqa-mistral-7B-v1 | 64.93% |\n| nv-embedqa-e5-v5 | 62.07% |\n| nv-embedqa-e5-v4 | 57.65% |\n| e5-large-unsupervised | 48.03% |\n| BM25 | 44.67% |\n\nWe evaluated the multilingual capabilities on the academic benchmark [MIRACL](https://github.com/project-miracl/miracl) across 15 languages and translated the English and Spanish version of MIRACL into additional 11 languages. The reported scores are based on an internal version of MIRACL by selecting hard negatives for each query to reduce the corpus size.\n\n| Open \u0026 Commercial Retrieval Models | Average Recall@5 on multilingual |\n| ----- | ----- |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 2048) | 60.75% |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 384) | 58.62% |\n| llama-3.2-nv-embedqa-1b-v1 | 60.07% |\n| nv-embedqa-mistral-7b-v2 | 50.42% |\n| BM25 | 26.51% |\n\nWe evaluated the cross-lingual capabilities on the academic benchmark [MLQA](https://github.com/facebookresearch/MLQA/) based on 7 languages (Arabic, Chinese, English, German, Hindi, Spanish, Vietnamese). We consider only evaluation datasets when the query and documents are in different languages. We calculate the average Recall@5 across the 42 different language pairs.\n\n| Open \u0026 Commercial Retrieval Models | Average Recall@5 on MLQA dataset with different languages |\n| ----- | ----- |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 2048) | 79.86% |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 384) | 71.61% |\n| llama-3.2-nv-embedqa-1b-v1 (embedding dim 2048) | 78.77% |\n| nv-embedqa-mistral-7b-v2 | 68.38% |\n| BM25 | 13.01% |\n\nWe evaluated the support of long documents on the academic benchmark [Multilingual Long-Document Retrieval (MLDR)](https://huggingface.co/datasets/Shitao/MLDR) built on Wikipedia and mC4, covering 12 typologically diverse languages. The English version has a median length of 2399 tokens and 90th percentile of 7483 tokens using the llama 3.2 tokenizer. The MLDR dataset is based on synthetic generated questions with a LLM, which has the tendency to create questions with similar keywords than the positive document, but might not be representative for real user queries. This characteristic of the dataset benefits sparse embeddings like BM25.\n\n| Open \u0026 Commercial Retrieval Models | Average Recall@5 on MLDR |\n| ----- | ----- |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 2048) | 59.55% |\n| llama-3.2-nv-embedqa-1b-v2 (embedding dim 384) | 54.77% |\n| llama-3.2-nv-embedqa-1b-v1 (embedding dim 2048) | 60.49% |\n| nv-embedqa-mistral-7b-v2 | 43.24% |\n| BM25 | 71.39% |\n\n**Data Collection Method by dataset**: Unknown\n\n**Labeling Method by dataset:** Unknown\n\n**Properties:** The evaluation datasets are based on [MTEB/BEIR](https://github.com/beir-cellar/beir), TextQA, TechQA, [MIRACL](https://github.com/project-miracl/miracl), [MLQA](https://github.com/facebookresearch/MLQA), and [MLDR](https://huggingface.co/datasets/Shitao/MLDR). The size ranges between 10,000s up to 5M depending on the dataset.\n\n**Inference**\n**Engine:** TensorRT\n**Test Hardware:** H100 PCIe/SXM, A100 PCIe/SXM, L40s, L4, and A10G\n\n## **Ethical Considerations**\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nFor more detailed information on ethical considerations for this model, please see the Model Card++ tab for the Explainability, Bias, Safety \u0026 Security, and Privacy subcards.\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"b9:T5cd,| Field | Response |\n| ----- | ----- |\n| Intended Application \u0026 Domain: | Passage and query embedding for question and answer retrieval |\n| Model Type: | Transformer encoder |\n| Intended User: | Generative AI creators working with conversational AI models - users who want to build a multilingual question and answer application over a large text corpus, leveraging the latest dense retrieval technologies. |\n| Output: | Array of float numbers (Dense Vector Representation for the input text) |\n| Describe how the model works: | Model transforms the tokenized input text into a dense vector representation. |\n| Performance Metrics: | Accuracy, Throughput, and Latency |\n| Potential Known Risks: | This model does not always guarantee to retrieve the correct passage(s) for a given query. |\n| Licensing \u0026 Terms of Use: | The use of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/) and Llama 3.2 is licensed under the [Llama 3.2 Community License](https://www.llama.com/llama3_2/license/), Copyright © Meta Platforms, Inc. All Rights Reserved. |\n| Technical Limitations | The model’s max sequence length is 8192. Therefore, the longer text inputs should be truncated. |\n| Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of: | N/A |\n| Verified to have met prescribed NVIDIA quality standards: | Yes |ba:T467,| Field | Response |\n| ----- | ----- |\n| Generatable or reverse engineerable personally-identifiable information (PII)? | None |\n| Was consent obtained for any personal data used? | Not Applicable |\n| PII used to create this model? | None |\n| How often is the dataset reviewed? | Before Every Release |\n| Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n| If personal data was collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable |\n| If personal data w"])</script><script>self.__next_f.push([1,"as collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable |\n| If personal data was collected for the development of this AI model, was it minimized to only what was required? | Not Applicable |\n| Is there provenance for all datasets used in training? | Yes |\n| Does data labeling (annotation, metadata) comply with privacy laws? | Yes |\n| Is data compliant with data subject requests for data correction or removal, if such a request was made? | No, not possible with externally-sourced data. |bb:Te384,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [\n -0.021759033203125,\n 0.028717041015625,\n -0.00675201416015625,\n 0.0290374755859375,\n 0.03314208984375,\n -0.00557708740234375,\n -0.0633544921875,\n -0.0167694091796875,\n -0.0277252197265625,\n -0.00653839111328125,\n -0.0016307830810546875,\n -0.004207611083984375,\n -0.0019245147705078125,\n -0.0130462646484375,\n -0.0254669189453125,\n 0.021575927734375,\n 0.0161590576171875,\n 0.018585205078125,\n 0.007266998291015625,\n 0.00820159912109375,\n -0.03369140625,\n -0.02069091796875,\n -0.023834228515625,\n 0.040008544921875,\n -0.0447998046875,\n -0.0032749176025390625,\n -0.04034423828125,\n -0.04119873046875,\n 0.0352783203125,\n -0.00421905517578125,\n 0.0225067138671875,\n 0.0217437744140625,\n -0.0171661376953125,\n 0.04888916015625,\n -0.0338134765625,\n 0.01395416259765625,\n -0.00406646728515625,\n -0.0194244384765625,\n 0.034027099609375,\n -0.00042819976806640625,\n -0.039031982421875,\n -0.01052093505859375,\n 0.014739990234375,\n -0.02410888671875,\n -0.01035308837890625,\n -0.01081085205078125,\n -0.0213775634765625,\n 0.027313232421875,\n -0.0224456787109375,\n 0.00890350341796875,\n 0.01194000244140625,\n 0.00978851318359375,\n -0.0150299072265625,\n 0.0024585723876953125,\n -0.006031036376953125,\n 0.025970458984375,\n -0.043975830078125,\n -0.0302581787109375,\n 0.024749755859375,\n -0.02191162109375,\n 0.0209503173828125,\n -0.016265869140625,\n -0.0068817138671875,\n 0.0233306884765625,\n -0.043975830078125,\n 0.007045745849609375,\n 0.027587890625,\n 0.035125732421875,\n -0.0194244384765625,\n -0.0002613067626953125,\n 0.018096923828125,\n -0.0177459716796875,\n 0.0125274658203125,\n -0.0184173583984375,\n 0.0244293212890625,\n 0.0276947021484375,\n 0.0321044921875,\n -0.004940032958984375,\n 0.02178955078125,\n 0.0218505859375,\n -0.038970947265625,\n -0.0095367431640625,\n -0.0325927734375,\n 0.02703857421875,\n -0.0080413818359375,\n 0.0192413330078125,\n 0.009124755859375,\n 0.012908935546875,\n 0.0093841552734375,\n -0.0633544921875,\n 0.024017333984375,\n 0.0099945068359375,\n 0.005001068115234375,\n -0.005474090576171875,\n 0.00970458984375,\n -0.01316070556640625,\n -0.0001823902130126953,\n -0.0029468536376953125,\n -0.00757598876953125,\n -0.01995849609375,\n -0.01483154296875,\n -0.0088348388671875,\n 0.0268402099609375,\n 0.0020751953125,\n -0.0011320114135742188,\n 0.005115509033203125,\n 0.0277252197265625,\n 0.029327392578125,\n -0.04644775390625,\n 0.00582122802734375,\n 0.004726409912109375,\n -0.0330810546875,\n -0.01296234130859375,\n -0.01062774658203125,\n -0.0246124267578125,\n -0.018798828125,\n 0.003887176513671875,\n -0.0160064697265625,\n -0.031524658203125,\n -0.0389404296875,\n -0.0137939453125,\n 0.00878143310546875,\n -0.0136566162109375,\n 0.01087188720703125,\n -0.0054473876953125,\n 0.015167236328125,\n 0.069091796875,\n 0.032806396484375,\n -0.0202484130859375,\n -0.007801055908203125,\n -0.038818359375,\n 0.01494598388671875,\n -0.0172119140625,\n -0.0095977783203125,\n -0.0178680419921875,\n 0.027313232421875,\n -0.005817413330078125,\n -0.0076751708984375,\n 0.02264404296875,\n -0.0153961181640625,\n 0.03326416015625,\n -0.00498199462890625,\n -0.004512786865234375,\n -0.02972412109375,\n -0.0021915435791015625,\n -0.017181396484375,\n -0.0096282958984375,\n 0.020477294921875,\n 0.0230560302734375,\n -0.0155792236328125,\n 0.0200347900390625,\n 0.00986480712890625,\n 0.0305938720703125,\n 0.0279541015625,\n -0.014495849609375,\n -0.0005102157592773438,\n 0.020599365234375,\n 0.02398681640625,\n 0.0012350082397460938,\n 0.02484130859375,\n 0.00826263427734375,\n -0.015289306640625,\n -0.02740478515625,\n -0.01605224609375,\n 0.0292816162109375,\n -0.01629638671875,\n 0.00946807861328125,\n -0.025970458984375,\n 0.0268707275390625,\n -0.0296630859375,\n -0.02618408203125,\n 0.0032749176025390625,\n -0.0163726806640625,\n 0.0517578125,\n 0.01922607421875,\n -0.0244903564453125,\n 0.007114410400390625,\n 0.00832366943359375,\n 0.01422882080078125,\n 0.024658203125,\n 0.01226806640625,\n -0.01342010498046875,\n -0.042083740234375,\n 0.00600433349609375,\n -0.041778564453125,\n 0.00902557373046875,\n 0.0028820037841796875,\n -0.040008544921875,\n 0.015960693359375,\n 0.039154052734375,\n 0.0144195556640625,\n 0.0135040283203125,\n 0.0180511474609375,\n -0.003978729248046875,\n 0.0174560546875,\n 0.005565643310546875,\n -0.01084136962890625,\n 0.00977325439453125,\n 0.01282501220703125,\n -0.0029811859130859375,\n 0.00212860107421875,\n 0.01934814453125,\n -0.02740478515625,\n -0.0242767333984375,\n -0.01226806640625,\n 0.0086669921875,\n -0.02459716796875,\n -0.007549285888671875,\n 0.006015777587890625,\n 0.0072784423828125,\n 0.016082763671875,\n 0.0167236328125,\n 0.00392913818359375,\n 0.0103912353515625,\n -0.039581298828125,\n 0.01537322998046875,\n 0.0015125274658203125,\n 0.00022912025451660156,\n 0.00824737548828125,\n -0.0130157470703125,\n -0.039581298828125,\n -0.0469970703125,\n 0.060760498046875,\n 0.0184173583984375,\n 0.04205322265625,\n 0.0035953521728515625,\n 0.009979248046875,\n 0.025848388671875,\n -0.033905029296875,\n -0.0222015380859375,\n 0.0003273487091064453,\n -0.0010213851928710938,\n 0.0014104843139648438,\n 0.0289459228515625,\n -0.03765869140625,\n -0.01273345947265625,\n 0.00919342041015625,\n 0.01200103759765625,\n -0.005420684814453125,\n 0.0175018310546875,\n 0.00031948089599609375,\n 0.0025348663330078125,\n -0.0174102783203125,\n -0.01151275634765625,\n -0.0091094970703125,\n 0.0192718505859375,\n 0.0181427001953125,\n 0.03228759765625,\n 0.03369140625,\n -0.018218994140625,\n -0.0018100738525390625,\n -0.018463134765625,\n 0.017974853515625,\n 0.0219573974609375,\n -0.019012451171875,\n 0.0034542083740234375,\n -0.034576416015625,\n 0.0386962890625,\n 0.006740570068359375,\n 0.0396728515625,\n 0.0123291015625,\n 0.03057861328125,\n 0.0043792724609375,\n -0.00820159912109375,\n -0.041412353515625,\n 0.0290374755859375,\n -0.0254364013671875,\n -0.00982666015625,\n 0.0078582763671875,\n 0.0117034912109375,\n -0.0264739990234375,\n -0.0198822021484375,\n 0.025970458984375,\n 0.005207061767578125,\n 0.01641845703125,\n 0.00011342763900756836,\n -0.021759033203125,\n 0.044891357421875,\n -0.012603759765625,\n -0.02178955078125,\n 0.0006842613220214844,\n 0.0242919921875,\n -0.0108184814453125,\n -0.018096923828125,\n 0.012359619140625,\n 0.013519287109375,\n -0.03619384765625,\n 0.0283203125,\n 0.0024852752685546875,\n 0.0092315673828125,\n 0.0015001296997070312,\n -0.0038700103759765625,\n 0.032958984375,\n -0.023223876953125,\n 0.01412200927734375,\n -0.024993896484375,\n 0.0022907257080078125,\n -0.03265380859375,\n -0.0118865966796875,\n -0.005062103271484375,\n -0.0198822021484375,\n 0.00820159912109375,\n 0.00240325927734375,\n -0.00860595703125,\n -0.01248931884765625,\n -0.004055023193359375,\n -0.051422119140625,\n 0.022918701171875,\n 0.019287109375,\n 0.0023670196533203125,\n -0.01224517822265625,\n -0.037750244140625,\n -0.033538818359375,\n -0.003391265869140625,\n 0.007801055908203125,\n 0.037384033203125,\n -0.011749267578125,\n -0.0266265869140625,\n -0.0230255126953125,\n -0.01200103759765625,\n -0.01751708984375,\n -0.0301513671875,\n 0.01092529296875,\n 0.0281219482421875,\n -0.004913330078125,\n 0.020538330078125,\n -0.044403076171875,\n 0.006710052490234375,\n 0.00205230712890625,\n 0.0246124267578125,\n 0.023895263671875,\n 0.003082275390625,\n -0.0183563232421875,\n -0.0025386810302734375,\n 0.0232696533203125,\n -0.026641845703125,\n 0.035430908203125,\n 0.00605010986328125,\n 0.004802703857421875,\n -0.0012407302856445312,\n -0.01342010498046875,\n -0.0276641845703125,\n 0.0075531005859375,\n 0.005832672119140625,\n -0.0117034912109375,\n 0.0186004638671875,\n 0.01314544677734375,\n -0.02557373046875,\n -0.0010271072387695312,\n 0.0047760009765625,\n -0.0245513916015625,\n 0.00025963783264160156,\n -0.0225067138671875,\n 0.007965087890625,\n -0.01361846923828125,\n -0.050323486328125,\n 0.035552978515625,\n 0.005481719970703125,\n -0.004589080810546875,\n 0.002834320068359375,\n -0.03271484375,\n -0.031951904296875,\n -0.0218505859375,\n 0.01320648193359375,\n 0.0225677490234375,\n -0.046234130859375,\n -0.004596710205078125,\n 0.023284912109375,\n -0.002410888671875,\n -0.00502777099609375,\n 0.01087188720703125,\n -0.01218414306640625,\n 0.0083465576171875,\n 0.006221771240234375,\n -0.004169464111328125,\n 0.01230621337890625,\n -0.0382080078125,\n -0.04205322265625,\n -0.00470733642578125,\n -0.0050201416015625,\n 0.0074310302734375,\n 0.0116119384765625,\n -0.00582122802734375,\n -0.00138092041015625,\n 0.01019287109375,\n -0.01309967041015625,\n 0.00902557373046875,\n 0.000263214111328125,\n -0.040008544921875,\n -0.02703857421875,\n -0.0271759033203125,\n -0.0330810546875,\n -0.0063018798828125,\n -0.049102783203125,\n -0.002166748046875,\n -0.034698486328125,\n -0.0230712890625,\n -0.0118865966796875,\n 0.0216827392578125,\n -0.004352569580078125,\n -0.0062103271484375,\n -0.015960693359375,\n -0.023223876953125,\n -0.0177764892578125,\n 0.0163726806640625,\n 0.01474761962890625,\n 0.01560211181640625,\n -0.00717926025390625,\n 0.03338623046875,\n 0.0391845703125,\n -0.020172119140625,\n 0.0025653839111328125,\n 0.0246124267578125,\n -0.0067291259765625,\n 0.0242919921875,\n -0.0169525146484375,\n -0.024383544921875,\n -0.014251708984375,\n -0.04412841796875,\n 0.01116943359375,\n 0.0048065185546875,\n 0.009033203125,\n 0.037811279296875,\n 0.0572509765625,\n 0.00787353515625,\n 0.0212249755859375,\n 0.023956298828125,\n 0.0242156982421875,\n 0.00580596923828125,\n 0.00852203369140625,\n -0.0009555816650390625,\n -0.0007157325744628906,\n 0.00832366943359375,\n 0.023956298828125,\n -0.016510009765625,\n -0.01372528076171875,\n -0.041290283203125,\n -0.020538330078125,\n -0.0252532958984375,\n 0.058319091796875,\n -0.0184173583984375,\n 0.01898193359375,\n -0.021453857421875,\n -0.00907135009765625,\n 0.0081634521484375,\n 0.0051422119140625,\n -0.0162353515625,\n 0.0147552490234375,\n -0.0012464523315429688,\n 0.0253753662109375,\n -0.01551055908203125,\n 0.043304443359375,\n 0.0214996337890625,\n 0.003818511962890625,\n 0.0309906005859375,\n 0.00994110107421875,\n 0.002166748046875,\n -0.01392364501953125,\n 0.01715087890625,\n -0.00824737548828125,\n -0.01666259765625,\n 0.04132080078125,\n 0.0178070068359375,\n 0.007373809814453125,\n 0.0129241943359375,\n 0.01381683349609375,\n 0.0052337646484375,\n 0.021209716796875,\n 0.0301055908203125,\n -0.0169219970703125,\n 0.00885009765625,\n -0.0185699462890625,\n 0.0430908203125,\n 0.01047515869140625,\n -0.00695037841796875,\n -0.039520263671875,\n 0.006504058837890625,\n 0.0031280517578125,\n -0.042144775390625,\n -0.004383087158203125,\n 0.01708984375,\n 0.040557861328125,\n 0.016693115234375,\n 0.0091400146484375,\n -0.02197265625,\n -0.0271759033203125,\n -0.00014901161193847656,\n -0.015899658203125,\n 0.010284423828125,\n -0.0135955810546875,\n 0.0186309814453125,\n -0.00966644287109375,\n 0.002201080322265625,\n -0.0249481201171875,\n -0.0254669189453125,\n 0.0183563232421875,\n -0.00933837890625,\n -0.00372314453125,\n 0.050872802734375,\n 0.0025482177734375,\n -0.05718994140625,\n -0.005062103271484375,\n 0.022918701171875,\n 0.02099609375,\n -0.04583740234375,\n -0.005626678466796875,\n 0.0011425018310546875,\n 0.03253173828125,\n 0.015045166015625,\n -0.0030155181884765625,\n 0.009613037109375,\n 0.030853271484375,\n 0.00821685791015625,\n 0.02093505859375,\n -0.0005106925964355469,\n 0.0027751922607421875,\n 0.022613525390625,\n -0.036346435546875,\n 0.0086212158203125,\n -0.004589080810546875,\n -0.030792236328125,\n -0.0095977783203125,\n -0.0196990966796875,\n 0.01247406005859375,\n -0.001712799072265625,\n 0.02252197265625,\n -0.0338134765625,\n -0.00850677490234375,\n -0.0218048095703125,\n -0.01611328125,\n 0.020172119140625,\n -0.0028133392333984375,\n 0.0004639625549316406,\n -0.01316070556640625,\n -0.028778076171875,\n -0.0194091796875,\n 0.01285552978515625,\n -0.005451202392578125,\n 0.0338134765625,\n 0.002105712890625,\n 0.0162200927734375,\n -0.0121002197265625,\n -0.0343017578125,\n -0.01480865478515625,\n 0.0005502700805664062,\n 0.0266876220703125,\n 0.01512908935546875,\n -0.0282745361328125,\n 0.0117034912109375,\n 0.0197906494140625,\n 0.0168304443359375,\n 0.016815185546875,\n -0.0101776123046875,\n -0.020599365234375,\n 0.030303955078125,\n 0.01113128662109375,\n -0.001064300537109375,\n 0.01007080078125,\n -0.017181396484375,\n 0.009918212890625,\n -0.004364013671875,\n -0.01041412353515625,\n -0.017669677734375,\n 0.04132080078125,\n 0.020416259765625,\n -0.00457000732421875,\n -0.0177764892578125,\n -0.01320648193359375,\n 0.0240936279296875,\n -0.049407958984375,\n -0.028778076171875,\n -0.017120361328125,\n -0.033843994140625,\n 0.00970458984375,\n -0.058868408203125,\n -0.0242767333984375,\n -0.00041365623474121094,\n 0.0037288665771484375,\n -0.01230621337890625,\n 0.039764404296875,\n 0.0152740478515625,\n 0.032806396484375,\n 0.011199951171875,\n -0.032379150390625,\n 0.053741455078125,\n 0.00446319580078125,\n 0.0865478515625,\n -0.010772705078125,\n -0.0032863616943359375,\n -0.042144775390625,\n 0.004840850830078125,\n -0.018218994140625,\n -0.0081634521484375,\n -0.005695343017578125,\n 0.00762939453125,\n -0.0282135009765625,\n 0.024200439453125,\n -0.036712646484375,\n -0.00927734375,\n -0.0007419586181640625,\n -0.008758544921875,\n -0.0195770263671875,\n -0.001636505126953125,\n -0.0118255615234375,\n 0.0238800048828125,\n -0.0019350051879882812,\n -0.00017178058624267578,\n -0.0210723876953125,\n 0.016357421875,\n 0.00824737548828125,\n -0.0014820098876953125,\n 0.044921875,\n 0.0233612060546875,\n 0.007740020751953125,\n -0.049560546875,\n -0.004245758056640625,\n 0.033050537109375,\n -0.005615234375,\n 0.02264404296875,\n -0.0087127685546875,\n 0.0223388671875,\n 0.01061248779296875,\n 0.00302886962890625,\n -0.0038318634033203125,\n 0.01471710205078125,\n -0.01548004150390625,\n 0.006481170654296875,\n 0.01554107666015625,\n -0.01403045654296875,\n 0.016571044921875,\n -0.024169921875,\n -0.019256591796875,\n -0.0010890960693359375,\n -0.0138702392578125,\n 0.0268402099609375,\n 0.01178741455078125,\n 0.0012445449829101562,\n 0.0126495361328125,\n -0.004123687744140625,\n 0.05645751953125,\n 0.0186004638671875,\n -0.0660400390625,\n -0.048980712890625,\n 0.0116424560546875,\n -0.035858154296875,\n 0.004688262939453125,\n -0.01494598388671875,\n 0.03009033203125,\n 0.007114410400390625,\n 0.004520416259765625,\n -0.016845703125,\n 0.01549530029296875,\n 0.0078582763671875,\n -0.061187744140625,\n -0.00006115436553955078,\n 0.02374267578125,\n -0.007205963134765625,\n -0.023284912109375,\n -0.01001739501953125,\n 0.0117340087890625,\n -0.01067352294921875,\n -0.005886077880859375,\n -0.01293182373046875,\n -0.0178680419921875,\n 0.00403594970703125,\n -0.0009860992431640625,\n -0.027862548828125,\n 0.027587890625,\n 0.0006814002990722656,\n 0.047210693359375,\n -0.00902557373046875,\n -0.029144287109375,\n 0.00928497314453125,\n -0.004913330078125,\n 0.02557373046875,\n 0.00682830810546875,\n -0.01922607421875,\n -0.002452850341796875,\n -0.0266876220703125,\n 0.01000213623046875,\n -0.01494598388671875,\n 0.01554107666015625,\n 0.005794525146484375,\n -0.031829833984375,\n -0.0293731689453125,\n 0.01239013671875,\n 0.004276275634765625,\n 0.003032684326171875,\n 0.0150146484375,\n 0.00402069091796875,\n -0.01136016845703125,\n 0.006481170654296875,\n -0.01727294921875,\n -0.0292205810546875,\n -0.0304412841796875,\n -0.020965576171875,\n -0.00982666015625,\n -0.0179901123046875,\n -0.035186767578125,\n -0.0214996337890625,\n 0.03265380859375,\n 0.023345947265625,\n -0.025482177734375,\n -0.005985260009765625,\n 0.0023441314697265625,\n 0.006561279296875,\n 0.00400543212890625,\n 0.032958984375,\n -0.00566864013671875,\n 0.0191497802734375,\n 0.00011974573135375977,\n 0.01081085205078125,\n -0.0224609375,\n -0.0294342041015625,\n 0.03863525390625,\n -0.016326904296875,\n 0.03204345703125,\n 0.013916015625,\n 0.0017852783203125,\n 0.0191650390625,\n -0.0242767333984375,\n -0.00342559814453125,\n -0.0218963623046875,\n 0.019378662109375,\n -0.011199951171875,\n 0.009307861328125,\n -0.0198974609375,\n -0.00701904296875,\n -0.01364898681640625,\n -0.0290374755859375,\n -0.00849151611328125,\n 0.004833221435546875,\n -0.00485992431640625,\n -0.0012979507446289062,\n 0.0251007080078125,\n 0.0169525146484375,\n -0.00701904296875,\n -0.0051116943359375,\n -0.0169830322265625,\n 0.04351806640625,\n -0.041412353515625,\n 0.00689697265625,\n -0.020782470703125,\n -0.01416015625,\n -0.033599853515625,\n 0.0013570785522460938,\n -0.04241943359375,\n -0.027191162109375,\n -0.01019287109375,\n -0.0164337158203125,\n 0.0165557861328125,\n -0.01922607421875,\n 0.036651611328125,\n -0.0261993408203125,\n -0.0012950897216796875,\n 0.02484130859375,\n 0.03759765625,\n -0.00919342041015625,\n 0.0182952880859375,\n 0.0310821533203125,\n 0.0141754150390625,\n -0.015533447265625,\n -0.03033447265625,\n -0.021575927734375,\n -0.02410888671875,\n 0.00817108154296875,\n 0.0011730194091796875,\n -0.0014801025390625,\n -0.0079803466796875,\n 0.0198822021484375,\n -0.0037078857421875,\n 0.0474853515625,\n 0.035736083984375,\n 0.012115478515625,\n 0.033538818359375,\n -0.0001016855239868164,\n -0.02001953125,\n 0.01580810546875,\n 0.0146026611328125,\n -0.025970458984375,\n 0.004581451416015625,\n -0.005603790283203125,\n 0.04888916015625,\n 0.00835418701171875,\n 0.01727294921875,\n 0.0167694091796875,\n 0.021697998046875,\n 0.0158233642578125,\n -0.023773193359375,\n 0.0262908935546875,\n -0.013763427734375,\n -0.0123291015625,\n -0.01165008544921875,\n 0.0171966552734375,\n -0.06005859375,\n 0.00960540771484375,\n 0.0177154541015625,\n 0.008514404296875,\n 0.0167388916015625,\n -0.01123046875,\n -0.035003662109375,\n 0.0006346702575683594,\n 0.004352569580078125,\n -0.041778564453125,\n 0.025665283203125,\n -0.0308074951171875,\n -0.034271240234375,\n 0.0043487548828125,\n -0.00344085693359375,\n 0.006683349609375,\n 0.0484619140625,\n -0.01386260986328125,\n -0.0052032470703125,\n -0.005847930908203125,\n 0.012298583984375,\n 0.022247314453125,\n 0.0404052734375,\n 0.0068511962890625,\n 0.0027618408203125,\n 0.0280609130859375,\n -0.0023174285888671875,\n 0.02099609375,\n -0.008544921875,\n 0.0016622543334960938,\n -0.0293121337890625,\n 0.023223876953125,\n -0.024932861328125,\n -0.003261566162109375,\n -0.007568359375,\n -0.0011339187622070312,\n 0.02117919921875,\n -0.055999755859375,\n 0.01334381103515625,\n 0.0017671585083007812,\n 0.0071258544921875,\n 0.00896453857421875,\n 0.0186004638671875,\n 0.007701873779296875,\n 0.0129241943359375,\n 0.0203094482421875,\n 0.0408935546875,\n 0.0239410400390625,\n -0.0037479400634765625,\n 0.01381683349609375,\n -0.00028634071350097656,\n -0.01364898681640625,\n -0.03131103515625,\n -0.00901031494140625,\n -0.0050506591796875,\n -0.0028743743896484375,\n -0.017822265625,\n -0.035980224609375,\n 0.05218505859375,\n 0.020965576171875,\n 0.0021915435791015625,\n 0.010772705078125,\n -0.01262664794921875,\n -0.01100921630859375,\n -0.0082550048828125,\n 0.0079345703125,\n 0.001178741455078125,\n 0.012939453125,\n 0.00982666015625,\n 0.01983642578125,\n 0.0214385986328125,\n -0.00847625732421875,\n -0.0152130126953125,\n 0.005268096923828125,\n -0.00836944580078125,\n -0.0160369873046875,\n 0.0235748291015625,\n -0.047149658203125,\n 0.0374755859375,\n 0.01532745361328125,\n 0.007053375244140625,\n -0.046356201171875,\n 0.04119873046875,\n -0.01507568359375,\n 0.029144287109375,\n 0.013702392578125,\n -0.0211029052734375,\n 0.007167816162109375,\n -0.02618408203125,\n -0.03326416015625,\n -0.0197296142578125,\n -0.0008974075317382812,\n 0.0299224853515625,\n 0.00274658203125,\n 0.028076171875,\n -0.0142059326171875,\n 0.03033447265625,\n 0.0036678314208984375,\n -0.0145721435546875,\n -0.0167236328125,\n -0.00594329833984375,\n 0.0191497802734375,\n 0.0277252197265625,\n 0.003032684326171875,\n 0.045928955078125,\n -0.0062103271484375,\n 0.0010519027709960938,\n -0.0185089111328125,\n 0.015838623046875,\n 0.02471923828125,\n -0.01451873779296875,\n 0.020538330078125,\n 0.00485992431640625,\n 0.0504150390625,\n -0.031494140625,\n 0.0243988037109375,\n -0.0189361572265625,\n 0.0063934326171875,\n 0.0287628173828125,\n -0.0008535385131835938,\n -0.0187225341796875,\n -0.0005669593811035156,\n 0.0293426513671875,\n 0.006412506103515625,\n -0.00036835670471191406,\n -0.013702392578125,\n 0.003551483154296875,\n -0.011322021484375,\n 0.002254486083984375,\n -0.01708984375,\n -0.0132598876953125,\n -0.0029506683349609375,\n -0.0272369384765625,\n -0.0286102294921875,\n 0.029541015625,\n 0.007965087890625,\n -0.002719879150390625,\n -0.00835418701171875,\n -0.0009479522705078125,\n 0.0094146728515625,\n 0.019287109375,\n -0.004444122314453125,\n 0.01206207275390625,\n 0.029327392578125,\n 0.006359100341796875,\n -0.021697998046875,\n 0.0050048828125,\n -0.0084381103515625,\n 0.02423095703125,\n -0.004180908203125,\n -0.01302337646484375,\n -0.03497314453125,\n -0.01154327392578125,\n -0.016632080078125,\n -0.0011310577392578125,\n 0.024658203125,\n 0.031646728515625,\n -0.034912109375,\n -0.0015211105346679688,\n -0.033538818359375,\n 0.0194549560546875,\n -0.01015472412109375,\n 0.024810791015625,\n 0.0208282470703125,\n -0.00013458728790283203,\n -0.03466796875,\n 0.0038318634033203125,\n 0.0260772705078125,\n 0.003482818603515625,\n -0.0226898193359375,\n -0.0224609375,\n 0.032989501953125,\n 0.035919189453125,\n -0.022064208984375,\n 0.059234619140625,\n -0.0006074905395507812,\n -0.002620697021484375,\n 0.00531768798828125,\n 0.0274658203125,\n 0.005001068115234375,\n -0.022308349609375,\n -0.004520416259765625,\n 0.016326904296875,\n -0.015045166015625,\n 0.02288818359375,\n -0.0012025833129882812,\n 0.0246124267578125,\n -0.0018339157104492188,\n 0.0195159912109375,\n -0.01480865478515625,\n -0.0400390625,\n 0.0280609130859375,\n -0.0169830322265625,\n 0.02789306640625,\n -0.02679443359375,\n 0.012603759765625,\n 0.0200958251953125,\n -0.00168609619140625,\n 0.00876617431640625,\n -0.020355224609375,\n -0.044677734375,\n -0.0126495361328125,\n 0.006809234619140625,\n -0.012420654296875,\n 0.0279541015625,\n 0.0184173583984375,\n 0.05517578125,\n -0.0240020751953125,\n 0.042205810546875,\n 0.050628662109375,\n -0.034576416015625,\n 0.0014581680297851562,\n 0.033966064453125,\n 0.020782470703125,\n 0.003265380859375,\n -0.00975799560546875,\n 0.0268402099609375,\n -0.00138092041015625,\n -0.03369140625,\n 0.01226806640625,\n -0.01471710205078125,\n 0.00582122802734375,\n -0.00021183490753173828,\n 0.0147705078125,\n -0.00708770751953125,\n -0.0310821533203125,\n -0.01053619384765625,\n -0.01145172119140625,\n -0.0140228271484375,\n 0.04461669921875,\n -0.034820556640625,\n 0.011566162109375,\n -0.031982421875,\n 0.0033721923828125,\n 0.00960540771484375,\n 0.0015716552734375,\n 0.02728271484375,\n -0.01507568359375,\n 0.0304412841796875,\n 0.04083251953125,\n 0.0002872943878173828,\n -0.040252685546875,\n 0.007045745849609375,\n 0.005306243896484375,\n -0.006122589111328125,\n -0.00746917724609375,\n -0.0008921623229980469,\n 0.01529693603515625,\n -0.01029205322265625,\n 0.0161285400390625,\n -0.003849029541015625,\n -0.018280029296875,\n -0.00005936622619628906,\n 0.024993896484375,\n 0.01430511474609375,\n -0.050994873046875,\n 0.010223388671875,\n -0.0006861686706542969,\n -0.00809478759765625,\n -0.006824493408203125,\n 0.004261016845703125,\n -0.01491546630859375,\n -0.007266998291015625,\n -0.0171966552734375,\n 0.003070831298828125,\n -0.0399169921875,\n 0.0191802978515625,\n 0.0013837814331054688,\n -0.013916015625,\n 0.01302337646484375,\n 0.000019550323486328125,\n 0.0027923583984375,\n 0.0247039794921875,\n -0.009429931640625,\n -0.0028705596923828125,\n 0.00457000732421875,\n -0.0142822265625,\n -0.01148223876953125,\n 0.0052490234375,\n -0.0159149169921875,\n -0.0019388198852539062,\n -0.037506103515625,\n 0.040008544921875,\n -0.01375579833984375,\n -0.0196990966796875,\n -0.0284881591796875,\n -0.045379638671875,\n -0.0181732177734375,\n 0.0174560546875,\n 0.004459381103515625,\n 0.01271820068359375,\n 0.01934814453125,\n 0.009796142578125,\n -0.005062103271484375,\n 0.0093841552734375,\n -0.0208892822265625,\n 0.0121917724609375,\n -0.0177764892578125,\n -0.006221771240234375,\n 0.03680419921875,\n 0.01007080078125,\n -0.029144287109375,\n -0.031341552734375,\n -0.0173492431640625,\n 0.0003707408905029297,\n -0.0161590576171875,\n -0.00705718994140625,\n 0.006793975830078125,\n 0.0013456344604492188,\n -0.0104217529296875,\n 0.033233642578125,\n -0.032318115234375,\n 0.02642822265625,\n -0.0277252197265625,\n -0.01290130615234375,\n 0.037750244140625,\n -0.00661468505859375,\n -0.0141754150390625,\n -0.007701873779296875,\n -0.01427459716796875,\n 0.0267486572265625,\n 0.0296173095703125,\n 0.020538330078125,\n -0.00782012939453125,\n -0.00708770751953125,\n -0.00970458984375,\n 0.0216827392578125,\n -0.002849578857421875,\n -0.01108551025390625,\n 0.0034046173095703125,\n 0.03961181640625,\n -0.0018873214721679688,\n -0.04412841796875,\n 0.00012421607971191406,\n -0.034912109375,\n 0.0191650390625,\n 0.0181884765625,\n 0.0227508544921875,\n -0.005626678466796875,\n -0.01312255859375,\n 0.0309600830078125,\n -0.0082550048828125,\n -0.025146484375,\n 0.0254364013671875,\n 0.0180816650390625,\n 0.01177215576171875,\n -0.034423828125,\n -0.01605224609375,\n 0.0127105712890625,\n 0.004840850830078125,\n -0.0209503173828125,\n 0.0396728515625,\n 0.0267333984375,\n 0.037261962890625,\n -0.003734588623046875,\n 0.020263671875,\n -0.038604736328125,\n -0.0745849609375,\n 0.037811279296875,\n -0.006977081298828125,\n 0.01464080810546875,\n -0.00995635986328125,\n -0.0208587646484375,\n -0.0273590087890625,\n 0.007343292236328125,\n -0.0005130767822265625,\n -0.00595855712890625,\n -0.0200042724609375,\n -0.025604248046875,\n -0.003459930419921875,\n 0.0141754150390625,\n 0.002593994140625,\n 0.0340576171875,\n 0.015167236328125,\n 0.002628326416015625,\n -0.0027828216552734375,\n -0.0104827880859375,\n 0.033843994140625,\n 0.028045654296875,\n 0.019622802734375,\n 0.0241241455078125,\n 0.01245880126953125,\n 0.0047607421875,\n -0.01715087890625,\n 0.02154541015625,\n 0.01226806640625,\n -0.0106353759765625,\n 0.056640625,\n 0.0219879150390625,\n 0.0202789306640625,\n -0.00531768798828125,\n -0.0143585205078125,\n -0.00009465217590332031,\n -0.01409149169921875,\n 0.013519287109375,\n 0.007480621337890625,\n 0.00637054443359375,\n -0.04168701171875,\n 0.01230621337890625,\n 0.037017822265625,\n -0.030548095703125,\n -0.01399993896484375,\n -0.006072998046875,\n -0.0183258056640625,\n -0.0284881591796875,\n -0.02154541015625,\n 0.004627227783203125,\n -0.01149749755859375,\n -0.019775390625,\n 0.037261962890625,\n -0.01158905029296875,\n -0.004848480224609375,\n -0.0293426513671875,\n -0.004528045654296875,\n -0.0015659332275390625,\n -0.020111083984375,\n 0.020111083984375,\n -0.021148681640625,\n -0.036773681640625,\n 0.004878997802734375,\n 0.0130767822265625,\n -0.002410888671875,\n -0.015960693359375,\n 0.037353515625,\n -0.008148193359375,\n -0.06597900390625,\n 0.03448486328125,\n -0.006744384765625,\n 0.01416015625,\n -0.0267333984375,\n 0.02886962890625,\n 0.0133209228515625,\n -0.0369873046875,\n -0.039459228515625,\n 0.028717041015625,\n 0.0205230712890625,\n 0.038330078125,\n -0.01535797119140625,\n 0.027587890625,\n 0.023223876953125,\n 0.003604888916015625,\n -0.0195465087890625,\n 0.01029205322265625,\n 0.004791259765625,\n 0.017059326171875,\n -0.0101165771484375,\n -0.035125732421875,\n 0.03857421875,\n 0.00849151611328125,\n 0.048675537109375,\n 0.0141754150390625,\n 0.0080413818359375,\n 0.0063018798828125,\n 0.006786346435546875,\n -0.0256500244140625,\n 0.0132904052734375,\n -0.007110595703125,\n -0.023406982421875,\n -0.049224853515625,\n -0.029327392578125,\n -0.0132598876953125,\n -0.00830078125,\n -0.00479888916015625,\n 0.0306396484375,\n -0.024383544921875,\n 0.0174560546875,\n -0.0092315673828125,\n -0.0196990966796875,\n -0.036376953125,\n -0.0027618408203125,\n 0.0029201507568359375,\n 0.001956939697265625,\n -0.0013761520385742188,\n 0.044708251953125,\n 0.00916290283203125,\n 0.0122833251953125,\n 0.007659912109375,\n 0.01381683349609375,\n -0.0430908203125,\n 0.0170135498046875,\n 0.028717041015625,\n -0.00998687744140625,\n 0.00720977783203125,\n 0.027740478515625,\n -0.01158905029296875,\n 0.0310821533203125,\n -0.01019287109375,\n 0.0019464492797851562,\n -0.033050537109375,\n 0.01126861572265625,\n 0.0130462646484375,\n 0.01861572265625,\n 0.007335662841796875,\n 0.027801513671875,\n 0.00594329833984375,\n 0.0122222900390625,\n 0.007232666015625,\n -0.033905029296875,\n 0.0274658203125,\n 0.001674652099609375,\n -0.027191162109375,\n -0.001514434814453125,\n -0.04278564453125,\n -0.0177001953125,\n 0.048187255859375,\n -0.03070068359375,\n -0.02191162109375,\n 0.031585693359375,\n 0.028900146484375,\n -0.00907135009765625,\n -0.020477294921875,\n 0.0037555694580078125,\n -0.0008893013000488281,\n -0.01071929931640625,\n 0.0199127197265625,\n -0.044403076171875,\n -0.049102783203125,\n -0.0143585205078125,\n -0.019073486328125,\n -0.0227508544921875,\n 0.0302734375,\n -0.00743865966796875,\n 0.005764007568359375,\n 0.00872802734375,\n 0.013519287109375,\n -0.0166168212890625,\n 0.0008749961853027344,\n -0.028076171875,\n 0.031280517578125,\n -0.002857208251953125,\n -0.006481170654296875,\n -0.0023212432861328125,\n -0.01312255859375,\n -0.02203369140625,\n -0.005268096923828125,\n -0.0103607177734375,\n -0.051422119140625,\n -0.0018215179443359375,\n 0.03460693359375,\n 0.0140533447265625,\n 0.00910186767578125,\n -0.011444091796875,\n -0.0007572174072265625,\n 0.0173492431640625,\n 0.01410675048828125,\n 0.016082763671875,\n 0.0048675537109375,\n -0.0283966064453125,\n 0.019622802734375,\n -0.01019287109375,\n 0.019805908203125,\n -0.0606689453125,\n -0.00238037109375,\n -0.0231170654296875,\n -0.00975799560546875,\n 0.010589599609375,\n 0.0298309326171875,\n -0.02703857421875,\n 0.03955078125,\n 0.0193634033203125,\n -0.021453857421875,\n 0.00897979736328125,\n -0.00994110107421875,\n 0.029937744140625,\n 0.00797271728515625,\n -0.08343505859375,\n -0.019073486328125,\n 0.0016622543334960938,\n -0.0122222900390625,\n -0.028106689453125,\n -0.0204925537109375,\n 0.01290130615234375,\n 0.047088623046875,\n 0.035858154296875,\n -0.006786346435546875,\n 0.01134490966796875,\n 0.02703857421875,\n -0.007747650146484375,\n 0.00020372867584228516,\n -0.0120849609375,\n -0.00899505615234375,\n 0.0016374588012695312,\n 0.01509857177734375,\n -0.01056671142578125,\n 0.0178375244140625,\n -0.03466796875,\n 0.032318115234375,\n -0.013763427734375,\n -0.0212554931640625,\n 0.00748443603515625,\n 0.0274658203125,\n -0.0164031982421875,\n -0.0208282470703125,\n -0.026458740234375,\n 0.0279541015625,\n -0.01727294921875,\n -0.0235137939453125,\n 0.009552001953125,\n 0.00211334228515625,\n -0.00917816162109375,\n 0.033050537109375,\n 0.005901336669921875,\n 0.0107574462890625,\n -0.01245880126953125,\n -0.00200653076171875,\n 0.0179901123046875,\n -0.005886077880859375,\n -0.01554107666015625,\n -0.034027099609375,\n 0.0009660720825195312,\n 0.0111236572265625,\n -0.0152587890625,\n 0.020721435546875,\n 0.028961181640625,\n -0.0043487548828125,\n -0.0024852752685546875,\n -0.0218505859375,\n 0.0007123947143554688,\n -0.0426025390625,\n -0.0006699562072753906,\n -0.00372314453125,\n -0.0248870849609375,\n -0.03167724609375,\n 0.0140533447265625,\n -0.01213836669921875,\n -0.00481414794921875,\n -0.0262451171875,\n -0.000385284423828125,\n -0.006072998046875,\n -0.004093170166015625,\n -0.06793212890625,\n -0.05712890625,\n -0.021209716796875,\n -0.0140533447265625,\n -0.03466796875,\n -0.0144500732421875,\n 0.01158905029296875,\n -0.0310211181640625,\n 0.03692626953125,\n -0.02899169921875,\n -0.00197601318359375,\n -0.0003044605255126953,\n 0.0013093948364257812,\n 0.0297088623046875,\n -0.00620269775390625,\n 0.005146026611328125,\n -0.01111602783203125,\n -0.00632476806640625,\n -0.02740478515625,\n -0.03729248046875,\n 0.0017337799072265625,\n 0.00995635986328125,\n -0.026458740234375,\n -0.006153106689453125,\n -0.00337982177734375,\n 0.018096923828125,\n 0.0384521484375,\n -0.00411224365234375,\n -0.033905029296875,\n -0.00334930419921875,\n -0.011505126953125,\n -0.020721435546875,\n -0.005702972412109375,\n -0.0042266845703125,\n 0.0078277587890625,\n 0.01265716552734375,\n -0.046356201171875,\n -0.0162200927734375,\n 0.007526397705078125,\n -0.015960693359375,\n -0.035552978515625,\n -0.003742218017578125,\n -0.005130767822265625,\n 0.020660400390625,\n -0.00572967529296875,\n -0.0260162353515625,\n 0.01235198974609375,\n 0.00004369020462036133,\n -0.0025691986083984375,\n 0.004543304443359375,\n 0.027587890625,\n -0.02783203125,\n -0.00531005859375,\n -0.007358551025390625,\n -0.0205841064453125,\n -0.044677734375,\n -0.00797271728515625,\n 0.006359100341796875,\n -0.0227508544921875,\n 0.0068359375,\n -0.03009033203125,\n -0.026763916015625,\n 0.01629638671875,\n -0.04046630859375,\n 0.012115478515625,\n -0.026031494140625,\n -0.01427459716796875,\n 0.023223876953125,\n -0.00116729736328125,\n 0.01038360595703125,\n -0.01216888427734375,\n 0.022064208984375,\n -0.023712158203125,\n 0.0279541015625,\n -0.005550384521484375,\n -0.0253143310546875,\n 0.00556182861328125,\n -0.029205322265625,\n -0.013214111328125,\n -0.01055145263671875,\n -0.01328277587890625,\n 0.0128021240234375,\n 0.0364990234375,\n 0.0180511474609375,\n -0.01708984375,\n 0.03228759765625,\n 0.0110626220703125,\n 0.00887298583984375,\n 0.01409149169921875,\n 0.005474090576171875,\n -0.00045371055603027344,\n -0.01168060302734375,\n 0.0130767822265625,\n -0.016632080078125,\n 0.0020599365234375,\n 0.022125244140625,\n -0.0035610198974609375,\n -0.0015239715576171875,\n 0.015655517578125,\n -0.037933349609375,\n 0.0235748291015625,\n -0.00960540771484375,\n -0.004741668701171875,\n 0.016265869140625,\n -0.005481719970703125,\n 0.01361846923828125,\n -0.02520751953125,\n -0.004581451416015625,\n 0.033111572265625,\n 0.0127410888671875,\n 0.02789306640625,\n -0.020172119140625,\n -0.013092041015625,\n 0.047454833984375,\n 0.0007686614990234375,\n -0.007740020751953125,\n 0.01328277587890625,\n 0.00412750244140625,\n -0.042877197265625,\n 0.01065826416015625,\n -0.00838470458984375,\n -0.020965576171875,\n 0.0203094482421875,\n 0.0104827880859375,\n 0.04010009765625,\n 0.021209716796875,\n -0.004917144775390625,\n -0.0108795166015625,\n 0.01244354248046875,\n 0.00846099853515625,\n 0.0088958740234375,\n -0.0037078857421875,\n -0.002422332763671875,\n -0.0440673828125,\n 0.0185089111328125,\n 0.004184722900390625,\n 0.048004150390625,\n -0.011993408203125,\n 0.0094451904296875,\n 0.005825042724609375,\n -0.0191497802734375,\n -0.0226287841796875,\n -0.0003876686096191406,\n -0.0022983551025390625,\n -0.00220489501953125,\n 0.00734710693359375,\n 0.0026340484619140625,\n 0.0244903564453125,\n -0.0172119140625,\n 0.0268096923828125,\n -0.034759521484375,\n 0.030029296875,\n 0.021453857421875,\n -0.020965576171875,\n 0.0146942138671875,\n -0.034210205078125,\n 0.01535797119140625,\n 0.034423828125,\n -0.0072479248046875,\n 0.00457000732421875,\n 0.04669189453125,\n -0.0341796875,\n 0.002040863037109375,\n -0.0019464492797851562,\n -0.0034046173095703125,\n 0.0020236968994140625,\n 0.0074310302734375,\n -0.01708984375,\n -0.037017822265625,\n -0.0146026611328125,\n -0.03546142578125,\n -0.0185089111328125,\n 0.01056671142578125,\n -0.040313720703125,\n -0.050506591796875,\n -0.00820159912109375,\n -0.0546875,\n -0.023681640625,\n -0.02484130859375,\n -0.0087127685546875,\n 0.03460693359375,\n 0.00009971857070922852,\n 0.0242919921875,\n -0.020416259765625,\n -0.0251007080078125,\n 0.0276641845703125,\n -0.008026123046875,\n -0.009521484375,\n 0.01690673828125,\n 0.000004410743713378906,\n -0.014251708984375,\n -0.002689361572265625,\n 0.011749267578125,\n 0.0197296142578125,\n -0.0276947021484375,\n 0.00514984130859375,\n 0.033905029296875,\n -0.0092926025390625,\n -0.0033512115478515625,\n -0.034088134765625,\n 0.023834228515625,\n -0.0019474029541015625,\n 0.00121307373046875,\n 0.019927978515625,\n -0.0287628173828125,\n -0.033538818359375,\n 0.006519317626953125,\n -0.0119476318359375,\n 0.0069427490234375,\n 0.050933837890625,\n -0.00904083251953125,\n -0.00496673583984375,\n 0.028961181640625,\n 0.00037407875061035156,\n 0.042266845703125,\n 0.0181732177734375,\n 0.0036163330078125,\n 0.01629638671875,\n 0.0158233642578125,\n 0.007030487060546875,\n -0.004718780517578125,\n 0.0017242431640625,\n -0.038909912109375,\n 0.0083465576171875,\n 0.00966644287109375,\n -0.004581451416015625,\n -0.0054931640625,\n 0.00978851318359375,\n 0.01027679443359375,\n -0.013702392578125,\n -0.0123748779296875,\n 0.0063934326171875,\n 0.027191162109375,\n -0.0309295654296875,\n -0.0217437744140625,\n -0.0249176025390625,\n 0.01073455810546875,\n -0.01218414306640625,\n -0.0238800048828125,\n 0.045501708984375,\n -0.028961181640625,\n 0.006290435791015625,\n 0.0095367431640625,\n -0.0235595703125,\n 0.005435943603515625,\n -0.036346435546875,\n -0.0163726806640625,\n -0.019012451171875,\n -0.0400390625,\n 0.0087127685546875,\n 0.01168060302734375,\n -0.019927978515625,\n -0.019317626953125,\n -0.0009331703186035156,\n -0.0004107952117919922,\n -0.0026073455810546875,\n 0.00342559814453125,\n 0.04571533203125,\n 0.007678985595703125,\n -0.0190887451171875,\n -0.01219940185546875,\n -0.020172119140625,\n -0.040771484375,\n 0.0132904052734375,\n 0.0028133392333984375,\n 0.006771087646484375,\n 0.004131317138671875,\n -0.0440673828125,\n 0.031585693359375,\n -0.0219879150390625,\n -0.026214599609375,\n 0.0290374755859375,\n -0.02520751953125,\n 0.0197601318359375,\n 0.02978515625,\n -0.015167236328125,\n -0.01180267333984375,\n -0.0316162109375,\n 0.001735687255859375,\n 0.01458740234375,\n -0.0009775161743164062,\n 0.0266265869140625,\n -0.0122222900390625,\n 0.0231170654296875,\n 0.012420654296875,\n 0.0240631103515625,\n -0.035980224609375,\n 0.03924560546875,\n -0.012420654296875,\n -0.0141754150390625,\n 0.02978515625,\n 0.01214599609375,\n 0.0256500244140625,\n -0.01507568359375,\n 0.0213775634765625,\n -0.00876617431640625,\n 0.0202789306640625,\n 0.0231170654296875,\n 0.0080413818359375,\n -0.0194854736328125,\n 0.00756072998046875,\n 0.0164031982421875,\n -0.0306549072265625,\n 0.0017375946044921875,\n 0.016357421875,\n -0.0054779052734375,\n 0.0240020751953125,\n -0.0023250579833984375,\n -0.01039886474609375,\n -0.0044097900390625,\n 0.0051727294921875,\n 0.0347900390625,\n 0.0203399658203125,\n -0.002277374267578125,\n -0.02691650390625,\n 0.01123046875,\n -0.0230712890625,\n 0.01221466064453125,\n -0.031951904296875,\n -0.0132904052734375,\n 0.0015573501586914062,\n 0.0081329345703125,\n -0.016326904296875,\n 0.00421142578125,\n 0.016845703125,\n 0.030548095703125,\n -0.0011539459228515625,\n -0.0196685791015625,\n -0.006153106689453125,\n 0.01003265380859375,\n -0.0030612945556640625,\n 0.03466796875,\n -0.01178741455078125,\n 0.0218658447265625,\n 0.01183319091796875,\n -0.015838623046875,\n -0.00782012939453125,\n -0.003528594970703125,\n -0.0225677490234375,\n 0.0199432373046875,\n -0.007427215576171875,\n 0.025604248046875,\n 0.0213470458984375,\n 0.0078582763671875,\n 0.01039886474609375,\n 0.021240234375,\n -0.012542724609375,\n 0.01611328125,\n 0.0313720703125,\n 0.00435638427734375,\n -0.0023021697998046875,\n -0.02069091796875,\n -0.0005769729614257812,\n 0.031585693359375,\n -0.005084991455078125,\n 0.0260162353515625,\n -0.012298583984375,\n -0.00618743896484375,\n 0.0129241943359375,\n -0.0419921875,\n -0.032318115234375,\n 0.01336669921875,\n 0.0343017578125,\n 0.037506103515625,\n -0.00975799560546875,\n 0.01068878173828125,\n -0.006450653076171875,\n -0.0036182403564453125,\n 0.0528564453125,\n 0.01364898681640625,\n 0.054290771484375,\n -0.041229248046875,\n 0.035980224609375,\n -0.01131439208984375,\n 0.022918701171875,\n 0.0023136138916015625,\n 0.041046142578125,\n -0.03228759765625,\n 0.033447265625,\n 0.0098114013671875,\n 0.01471710205078125,\n -0.01451873779296875,\n -0.003505706787109375,\n -0.0007944107055664062,\n 0.00888824462890625,\n -0.014434814453125,\n -0.01560211181640625,\n 0.039581298828125,\n -0.036712646484375,\n -0.01690673828125,\n 0.015777587890625,\n -0.00433349609375,\n -0.0199432373046875,\n -0.00594329833984375,\n -0.00966644287109375,\n -0.006195068359375,\n 0.0047454833984375,\n -0.0246124267578125,\n 0.01549530029296875,\n -0.0179290771484375,\n 0.053466796875,\n 0.051055908203125,\n 0.0242919921875,\n 0.0237884521484375,\n 0.01090240478515625,\n -0.02557373046875,\n -0.00540924072265625,\n 0.0299835205078125,\n 0.032684326171875,\n 0.0180511474609375,\n 0.0338134765625,\n 0.000301361083984375,\n -0.020538330078125,\n 0.0290374755859375,\n 0.01155853271484375,\n 0.05328369140625,\n -0.01016998291015625,\n 0.0048675537109375,\n 0.0265655517578125,\n 0.006988525390625,\n -0.02008056640625,\n -0.02459716796875,\n -0.04034423828125,\n -0.0001156926155090332,\n -0.017120361328125,\n -0.0312042236328125,\n -0.08050537109375,\n 0.009246826171875,\n -0.01177215576171875,\n 0.01433563232421875,\n -0.0005059242248535156,\n -0.01898193359375,\n 0.0146026611328125,\n 0.00020968914031982422,\n -0.0282440185546875,\n -0.020721435546875,\n -0.0078887939453125,\n 0.00934600830078125,\n 0.01219940185546875,\n 0.02130126953125,\n 0.036163330078125,\n -0.00826263427734375,\n 0.0126495361328125,\n 0.0166473388671875,\n 0.00372314453125,\n -0.0018815994262695312,\n 0.038543701171875,\n 0.0010766983032226562,\n -0.0098114013671875,\n -0.0132598876953125,\n -0.0217132568359375,\n -0.0137939453125,\n -0.00937652587890625,\n 0.028900146484375,\n -0.0271453857421875,\n -0.007259368896484375,\n -0.01959228515625,\n -0.00433349609375,\n -0.0205535888671875,\n -0.0183258056640625,\n -0.0127105712890625,\n 0.006977081298828125,\n 0.0018367767333984375,\n -0.0030155181884765625,\n 0.00959014892578125,\n -0.00818634033203125,\n 0.01108551025390625,\n -0.0304107666015625,\n -0.0114593505859375,\n 0.0219879150390625,\n -0.00087738037109375,\n -0.0239410400390625,\n 0.00443267822265625,\n -0.0012035369873046875,\n 0.034759521484375,\n 0.01416015625,\n -0.037567138671875,\n 0.031280517578125,\n -0.006862640380859375,\n -0.02508544921875,\n -0.0161895751953125,\n 0.02728271484375,\n -0.01139068603515625,\n 0.00475311279296875,\n 0.000667572021484375,\n 0.0225677490234375,\n -0.00963592529296875,\n 0.025482177734375,\n -0.018463134765625,\n -0.039581298828125,\n -0.0408935546875,\n -0.0138702392578125,\n 0.0260772705078125,\n 0.020965576171875,\n 0.030792236328125,\n 0.028167724609375,\n -0.0185546875,\n 0.004425048828125,\n -0.028594970703125,\n 0.00897979736328125,\n -0.01424407958984375,\n -0.0096893310546875,\n 0.0020999908447265625,\n -0.01512908935546875,\n 0.026458740234375,\n 0.005672454833984375,\n 0.0175628662109375,\n 0.005825042724609375,\n -0.0183563232421875,\n 0.0216827392578125,\n -0.029754638671875,\n -0.0222015380859375,\n -0.003185272216796875,\n 0.00878143310546875,\n -0.0152435302734375,\n 0.006740570068359375,\n 0.006793975830078125,\n -0.003658294677734375,\n 0.042510986328125,\n -0.02691650390625,\n 0.0237884521484375,\n -0.0021228790283203125,\n -0.004825592041015625,\n -0.017059326171875,\n 0.054901123046875,\n 0.0268096923828125,\n 0.04888916015625,\n 0.013641357421875,\n 0.02520751953125,\n 0.016021728515625,\n 0.00025916099548339844,\n 0.027130126953125,\n 0.034912109375,\n 0.0085296630859375,\n -0.0081329345703125,\n -0.00397491455078125,\n -0.01399993896484375,\n -0.0082244873046875,\n 0.01397705078125,\n -0.005687713623046875,\n -0.01282501220703125,\n 0.0172882080078125,\n 0.025146484375,\n -0.0218963623046875,\n -0.0019073486328125,\n 0.01097869873046875,\n 0.0016794204711914062,\n -0.002399444580078125,\n -0.0085296630859375,\n 0.0281829833984375,\n 0.0252532958984375,\n -0.0009751319885253906,\n -0.007537841796875,\n -0.020172119140625,\n -0.0159912109375,\n -0.0180511474609375,\n 0.0071258544921875,\n 0.004245758056640625,\n -0.0142822265625,\n -0.019012451171875,\n 0.0181884765625,\n -0.032958984375,\n -0.0211334228515625,\n -0.0255279541015625,\n -0.0191802978515625,\n -0.045989990234375,\n -0.00395965576171875,\n -0.004886627197265625,\n -0.0214996337890625,\n 0.02154541015625,\n 0.007537841796875,\n 0.055908203125,\n -0.01203155517578125,\n -0.0157623291015625,\n 0.00783538818359375,\n 0.012908935546875,\n -0.0175628662109375,\n -0.03204345703125,\n -0.01296234130859375,\n 0.03265380859375,\n 0.01934814453125,\n -0.00917816162109375,\n -0.0208892822265625,\n -0.0281219482421875,\n -0.0008893013000488281,\n -0.0077667236328125,\n -0.01224517822265625,\n -0.00033354759216308594,\n -0.006439208984375,\n -0.01094818115234375,\n -0.036956787109375,\n -0.023651123046875,\n -0.0230560302734375,\n 0.0165557861328125,\n 0.01922607421875,\n -0.008819580078125,\n -0.0121002197265625,\n 0.027313232421875,\n 0.007122039794921875,\n -0.033843994140625,\n -0.01560211181640625,\n -0.026397705078125,\n 0.0162200927734375,\n 0.01751708984375,\n 0.0229644775390625,\n -0.017822265625,\n 0.0267486572265625,\n 0.044403076171875,\n -0.018798828125,\n 0.03265380859375,\n 0.02349853515625,\n -0.041351318359375,\n -0.05206298828125,\n 0.01296234130859375,\n -0.0015277862548828125,\n -0.0010213851928710938,\n -0.0303802490234375,\n 0.0015411376953125,\n 0.00531005859375,\n 0.02239990234375,\n -0.007843017578125,\n -0.01262664794921875,\n 0.02899169921875,\n 0.036651611328125\n ],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/llama-3.2-nv-embedqa-1b-v2\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"bc:Te384,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [\n -0.021759033203125,\n 0.028717041015625,\n -0.00675201416015625,\n 0.0290374755859375,\n 0.03314208984375,\n -0.00557708740234375,\n -0.0633544921875,\n -0.0167694091796875,\n -0.0277252197265625,\n -0.00653839111328125,\n -0.0016307830810546875,\n -0.004207611083984375,\n -0.0019245147705078125,\n -0.0130462646484375,\n -0.0254669189453125,\n 0.021575927734375,\n 0.0161590576171875,\n 0.018585205078125,\n 0.007266998291015625,\n 0.00820159912109375,\n -0.03369140625,\n -0.02069091796875,\n -0.023834228515625,\n 0.040008544921875,\n -0.0447998046875,\n -0.0032749176025390625,\n -0.04034423828125,\n -0.04119873046875,\n 0.0352783203125,\n -0.00421905517578125,\n 0.0225067138671875,\n 0.0217437744140625,\n -0.0171661376953125,\n 0.04888916015625,\n -0.0338134765625,\n 0.01395416259765625,\n -0.00406646728515625,\n -0.0194244384765625,\n 0.034027099609375,\n -0.00042819976806640625,\n -0.039031982421875,\n -0.01052093505859375,\n 0.014739990234375,\n -0.02410888671875,\n -0.01035308837890625,\n -0.01081085205078125,\n -0.0213775634765625,\n 0.027313232421875,\n -0.0224456787109375,\n 0.00890350341796875,\n 0.01194000244140625,\n 0.00978851318359375,\n -0.0150299072265625,\n 0.0024585723876953125,\n -0.006031036376953125,\n 0.025970458984375,\n -0.043975830078125,\n -0.0302581787109375,\n 0.024749755859375,\n -0.02191162109375,\n 0.0209503173828125,\n -0.016265869140625,\n -0.0068817138671875,\n 0.0233306884765625,\n -0.043975830078125,\n 0.007045745849609375,\n 0.027587890625,\n 0.035125732421875,\n -0.0194244384765625,\n -0.0002613067626953125,\n 0.018096923828125,\n -0.0177459716796875,\n 0.0125274658203125,\n -0.0184173583984375,\n 0.0244293212890625,\n 0.0276947021484375,\n 0.0321044921875,\n -0.004940032958984375,\n 0.02178955078125,\n 0.0218505859375,\n -0.038970947265625,\n -0.0095367431640625,\n -0.0325927734375,\n 0.02703857421875,\n -0.0080413818359375,\n 0.0192413330078125,\n 0.009124755859375,\n 0.012908935546875,\n 0.0093841552734375,\n -0.0633544921875,\n 0.024017333984375,\n 0.0099945068359375,\n 0.005001068115234375,\n -0.005474090576171875,\n 0.00970458984375,\n -0.01316070556640625,\n -0.0001823902130126953,\n -0.0029468536376953125,\n -0.00757598876953125,\n -0.01995849609375,\n -0.01483154296875,\n -0.0088348388671875,\n 0.0268402099609375,\n 0.0020751953125,\n -0.0011320114135742188,\n 0.005115509033203125,\n 0.0277252197265625,\n 0.029327392578125,\n -0.04644775390625,\n 0.00582122802734375,\n 0.004726409912109375,\n -0.0330810546875,\n -0.01296234130859375,\n -0.01062774658203125,\n -0.0246124267578125,\n -0.018798828125,\n 0.003887176513671875,\n -0.0160064697265625,\n -0.031524658203125,\n -0.0389404296875,\n -0.0137939453125,\n 0.00878143310546875,\n -0.0136566162109375,\n 0.01087188720703125,\n -0.0054473876953125,\n 0.015167236328125,\n 0.069091796875,\n 0.032806396484375,\n -0.0202484130859375,\n -0.007801055908203125,\n -0.038818359375,\n 0.01494598388671875,\n -0.0172119140625,\n -0.0095977783203125,\n -0.0178680419921875,\n 0.027313232421875,\n -0.005817413330078125,\n -0.0076751708984375,\n 0.02264404296875,\n -0.0153961181640625,\n 0.03326416015625,\n -0.00498199462890625,\n -0.004512786865234375,\n -0.02972412109375,\n -0.0021915435791015625,\n -0.017181396484375,\n -0.0096282958984375,\n 0.020477294921875,\n 0.0230560302734375,\n -0.0155792236328125,\n 0.0200347900390625,\n 0.00986480712890625,\n 0.0305938720703125,\n 0.0279541015625,\n -0.014495849609375,\n -0.0005102157592773438,\n 0.020599365234375,\n 0.02398681640625,\n 0.0012350082397460938,\n 0.02484130859375,\n 0.00826263427734375,\n -0.015289306640625,\n -0.02740478515625,\n -0.01605224609375,\n 0.0292816162109375,\n -0.01629638671875,\n 0.00946807861328125,\n -0.025970458984375,\n 0.0268707275390625,\n -0.0296630859375,\n -0.02618408203125,\n 0.0032749176025390625,\n -0.0163726806640625,\n 0.0517578125,\n 0.01922607421875,\n -0.0244903564453125,\n 0.007114410400390625,\n 0.00832366943359375,\n 0.01422882080078125,\n 0.024658203125,\n 0.01226806640625,\n -0.01342010498046875,\n -0.042083740234375,\n 0.00600433349609375,\n -0.041778564453125,\n 0.00902557373046875,\n 0.0028820037841796875,\n -0.040008544921875,\n 0.015960693359375,\n 0.039154052734375,\n 0.0144195556640625,\n 0.0135040283203125,\n 0.0180511474609375,\n -0.003978729248046875,\n 0.0174560546875,\n 0.005565643310546875,\n -0.01084136962890625,\n 0.00977325439453125,\n 0.01282501220703125,\n -0.0029811859130859375,\n 0.00212860107421875,\n 0.01934814453125,\n -0.02740478515625,\n -0.0242767333984375,\n -0.01226806640625,\n 0.0086669921875,\n -0.02459716796875,\n -0.007549285888671875,\n 0.006015777587890625,\n 0.0072784423828125,\n 0.016082763671875,\n 0.0167236328125,\n 0.00392913818359375,\n 0.0103912353515625,\n -0.039581298828125,\n 0.01537322998046875,\n 0.0015125274658203125,\n 0.00022912025451660156,\n 0.00824737548828125,\n -0.0130157470703125,\n -0.039581298828125,\n -0.0469970703125,\n 0.060760498046875,\n 0.0184173583984375,\n 0.04205322265625,\n 0.0035953521728515625,\n 0.009979248046875,\n 0.025848388671875,\n -0.033905029296875,\n -0.0222015380859375,\n 0.0003273487091064453,\n -0.0010213851928710938,\n 0.0014104843139648438,\n 0.0289459228515625,\n -0.03765869140625,\n -0.01273345947265625,\n 0.00919342041015625,\n 0.01200103759765625,\n -0.005420684814453125,\n 0.0175018310546875,\n 0.00031948089599609375,\n 0.0025348663330078125,\n -0.0174102783203125,\n -0.01151275634765625,\n -0.0091094970703125,\n 0.0192718505859375,\n 0.0181427001953125,\n 0.03228759765625,\n 0.03369140625,\n -0.018218994140625,\n -0.0018100738525390625,\n -0.018463134765625,\n 0.017974853515625,\n 0.0219573974609375,\n -0.019012451171875,\n 0.0034542083740234375,\n -0.034576416015625,\n 0.0386962890625,\n 0.006740570068359375,\n 0.0396728515625,\n 0.0123291015625,\n 0.03057861328125,\n 0.0043792724609375,\n -0.00820159912109375,\n -0.041412353515625,\n 0.0290374755859375,\n -0.0254364013671875,\n -0.00982666015625,\n 0.0078582763671875,\n 0.0117034912109375,\n -0.0264739990234375,\n -0.0198822021484375,\n 0.025970458984375,\n 0.005207061767578125,\n 0.01641845703125,\n 0.00011342763900756836,\n -0.021759033203125,\n 0.044891357421875,\n -0.012603759765625,\n -0.02178955078125,\n 0.0006842613220214844,\n 0.0242919921875,\n -0.0108184814453125,\n -0.018096923828125,\n 0.012359619140625,\n 0.013519287109375,\n -0.03619384765625,\n 0.0283203125,\n 0.0024852752685546875,\n 0.0092315673828125,\n 0.0015001296997070312,\n -0.0038700103759765625,\n 0.032958984375,\n -0.023223876953125,\n 0.01412200927734375,\n -0.024993896484375,\n 0.0022907257080078125,\n -0.03265380859375,\n -0.0118865966796875,\n -0.005062103271484375,\n -0.0198822021484375,\n 0.00820159912109375,\n 0.00240325927734375,\n -0.00860595703125,\n -0.01248931884765625,\n -0.004055023193359375,\n -0.051422119140625,\n 0.022918701171875,\n 0.019287109375,\n 0.0023670196533203125,\n -0.01224517822265625,\n -0.037750244140625,\n -0.033538818359375,\n -0.003391265869140625,\n 0.007801055908203125,\n 0.037384033203125,\n -0.011749267578125,\n -0.0266265869140625,\n -0.0230255126953125,\n -0.01200103759765625,\n -0.01751708984375,\n -0.0301513671875,\n 0.01092529296875,\n 0.0281219482421875,\n -0.004913330078125,\n 0.020538330078125,\n -0.044403076171875,\n 0.006710052490234375,\n 0.00205230712890625,\n 0.0246124267578125,\n 0.023895263671875,\n 0.003082275390625,\n -0.0183563232421875,\n -0.0025386810302734375,\n 0.0232696533203125,\n -0.026641845703125,\n 0.035430908203125,\n 0.00605010986328125,\n 0.004802703857421875,\n -0.0012407302856445312,\n -0.01342010498046875,\n -0.0276641845703125,\n 0.0075531005859375,\n 0.005832672119140625,\n -0.0117034912109375,\n 0.0186004638671875,\n 0.01314544677734375,\n -0.02557373046875,\n -0.0010271072387695312,\n 0.0047760009765625,\n -0.0245513916015625,\n 0.00025963783264160156,\n -0.0225067138671875,\n 0.007965087890625,\n -0.01361846923828125,\n -0.050323486328125,\n 0.035552978515625,\n 0.005481719970703125,\n -0.004589080810546875,\n 0.002834320068359375,\n -0.03271484375,\n -0.031951904296875,\n -0.0218505859375,\n 0.01320648193359375,\n 0.0225677490234375,\n -0.046234130859375,\n -0.004596710205078125,\n 0.023284912109375,\n -0.002410888671875,\n -0.00502777099609375,\n 0.01087188720703125,\n -0.01218414306640625,\n 0.0083465576171875,\n 0.006221771240234375,\n -0.004169464111328125,\n 0.01230621337890625,\n -0.0382080078125,\n -0.04205322265625,\n -0.00470733642578125,\n -0.0050201416015625,\n 0.0074310302734375,\n 0.0116119384765625,\n -0.00582122802734375,\n -0.00138092041015625,\n 0.01019287109375,\n -0.01309967041015625,\n 0.00902557373046875,\n 0.000263214111328125,\n -0.040008544921875,\n -0.02703857421875,\n -0.0271759033203125,\n -0.0330810546875,\n -0.0063018798828125,\n -0.049102783203125,\n -0.002166748046875,\n -0.034698486328125,\n -0.0230712890625,\n -0.0118865966796875,\n 0.0216827392578125,\n -0.004352569580078125,\n -0.0062103271484375,\n -0.015960693359375,\n -0.023223876953125,\n -0.0177764892578125,\n 0.0163726806640625,\n 0.01474761962890625,\n 0.01560211181640625,\n -0.00717926025390625,\n 0.03338623046875,\n 0.0391845703125,\n -0.020172119140625,\n 0.0025653839111328125,\n 0.0246124267578125,\n -0.0067291259765625,\n 0.0242919921875,\n -0.0169525146484375,\n -0.024383544921875,\n -0.014251708984375,\n -0.04412841796875,\n 0.01116943359375,\n 0.0048065185546875,\n 0.009033203125,\n 0.037811279296875,\n 0.0572509765625,\n 0.00787353515625,\n 0.0212249755859375,\n 0.023956298828125,\n 0.0242156982421875,\n 0.00580596923828125,\n 0.00852203369140625,\n -0.0009555816650390625,\n -0.0007157325744628906,\n 0.00832366943359375,\n 0.023956298828125,\n -0.016510009765625,\n -0.01372528076171875,\n -0.041290283203125,\n -0.020538330078125,\n -0.0252532958984375,\n 0.058319091796875,\n -0.0184173583984375,\n 0.01898193359375,\n -0.021453857421875,\n -0.00907135009765625,\n 0.0081634521484375,\n 0.0051422119140625,\n -0.0162353515625,\n 0.0147552490234375,\n -0.0012464523315429688,\n 0.0253753662109375,\n -0.01551055908203125,\n 0.043304443359375,\n 0.0214996337890625,\n 0.003818511962890625,\n 0.0309906005859375,\n 0.00994110107421875,\n 0.002166748046875,\n -0.01392364501953125,\n 0.01715087890625,\n -0.00824737548828125,\n -0.01666259765625,\n 0.04132080078125,\n 0.0178070068359375,\n 0.007373809814453125,\n 0.0129241943359375,\n 0.01381683349609375,\n 0.0052337646484375,\n 0.021209716796875,\n 0.0301055908203125,\n -0.0169219970703125,\n 0.00885009765625,\n -0.0185699462890625,\n 0.0430908203125,\n 0.01047515869140625,\n -0.00695037841796875,\n -0.039520263671875,\n 0.006504058837890625,\n 0.0031280517578125,\n -0.042144775390625,\n -0.004383087158203125,\n 0.01708984375,\n 0.040557861328125,\n 0.016693115234375,\n 0.0091400146484375,\n -0.02197265625,\n -0.0271759033203125,\n -0.00014901161193847656,\n -0.015899658203125,\n 0.010284423828125,\n -0.0135955810546875,\n 0.0186309814453125,\n -0.00966644287109375,\n 0.002201080322265625,\n -0.0249481201171875,\n -0.0254669189453125,\n 0.0183563232421875,\n -0.00933837890625,\n -0.00372314453125,\n 0.050872802734375,\n 0.0025482177734375,\n -0.05718994140625,\n -0.005062103271484375,\n 0.022918701171875,\n 0.02099609375,\n -0.04583740234375,\n -0.005626678466796875,\n 0.0011425018310546875,\n 0.03253173828125,\n 0.015045166015625,\n -0.0030155181884765625,\n 0.009613037109375,\n 0.030853271484375,\n 0.00821685791015625,\n 0.02093505859375,\n -0.0005106925964355469,\n 0.0027751922607421875,\n 0.022613525390625,\n -0.036346435546875,\n 0.0086212158203125,\n -0.004589080810546875,\n -0.030792236328125,\n -0.0095977783203125,\n -0.0196990966796875,\n 0.01247406005859375,\n -0.001712799072265625,\n 0.02252197265625,\n -0.0338134765625,\n -0.00850677490234375,\n -0.0218048095703125,\n -0.01611328125,\n 0.020172119140625,\n -0.0028133392333984375,\n 0.0004639625549316406,\n -0.01316070556640625,\n -0.028778076171875,\n -0.0194091796875,\n 0.01285552978515625,\n -0.005451202392578125,\n 0.0338134765625,\n 0.002105712890625,\n 0.0162200927734375,\n -0.0121002197265625,\n -0.0343017578125,\n -0.01480865478515625,\n 0.0005502700805664062,\n 0.0266876220703125,\n 0.01512908935546875,\n -0.0282745361328125,\n 0.0117034912109375,\n 0.0197906494140625,\n 0.0168304443359375,\n 0.016815185546875,\n -0.0101776123046875,\n -0.020599365234375,\n 0.030303955078125,\n 0.01113128662109375,\n -0.001064300537109375,\n 0.01007080078125,\n -0.017181396484375,\n 0.009918212890625,\n -0.004364013671875,\n -0.01041412353515625,\n -0.017669677734375,\n 0.04132080078125,\n 0.020416259765625,\n -0.00457000732421875,\n -0.0177764892578125,\n -0.01320648193359375,\n 0.0240936279296875,\n -0.049407958984375,\n -0.028778076171875,\n -0.017120361328125,\n -0.033843994140625,\n 0.00970458984375,\n -0.058868408203125,\n -0.0242767333984375,\n -0.00041365623474121094,\n 0.0037288665771484375,\n -0.01230621337890625,\n 0.039764404296875,\n 0.0152740478515625,\n 0.032806396484375,\n 0.011199951171875,\n -0.032379150390625,\n 0.053741455078125,\n 0.00446319580078125,\n 0.0865478515625,\n -0.010772705078125,\n -0.0032863616943359375,\n -0.042144775390625,\n 0.004840850830078125,\n -0.018218994140625,\n -0.0081634521484375,\n -0.005695343017578125,\n 0.00762939453125,\n -0.0282135009765625,\n 0.024200439453125,\n -0.036712646484375,\n -0.00927734375,\n -0.0007419586181640625,\n -0.008758544921875,\n -0.0195770263671875,\n -0.001636505126953125,\n -0.0118255615234375,\n 0.0238800048828125,\n -0.0019350051879882812,\n -0.00017178058624267578,\n -0.0210723876953125,\n 0.016357421875,\n 0.00824737548828125,\n -0.0014820098876953125,\n 0.044921875,\n 0.0233612060546875,\n 0.007740020751953125,\n -0.049560546875,\n -0.004245758056640625,\n 0.033050537109375,\n -0.005615234375,\n 0.02264404296875,\n -0.0087127685546875,\n 0.0223388671875,\n 0.01061248779296875,\n 0.00302886962890625,\n -0.0038318634033203125,\n 0.01471710205078125,\n -0.01548004150390625,\n 0.006481170654296875,\n 0.01554107666015625,\n -0.01403045654296875,\n 0.016571044921875,\n -0.024169921875,\n -0.019256591796875,\n -0.0010890960693359375,\n -0.0138702392578125,\n 0.0268402099609375,\n 0.01178741455078125,\n 0.0012445449829101562,\n 0.0126495361328125,\n -0.004123687744140625,\n 0.05645751953125,\n 0.0186004638671875,\n -0.0660400390625,\n -0.048980712890625,\n 0.0116424560546875,\n -0.035858154296875,\n 0.004688262939453125,\n -0.01494598388671875,\n 0.03009033203125,\n 0.007114410400390625,\n 0.004520416259765625,\n -0.016845703125,\n 0.01549530029296875,\n 0.0078582763671875,\n -0.061187744140625,\n -0.00006115436553955078,\n 0.02374267578125,\n -0.007205963134765625,\n -0.023284912109375,\n -0.01001739501953125,\n 0.0117340087890625,\n -0.01067352294921875,\n -0.005886077880859375,\n -0.01293182373046875,\n -0.0178680419921875,\n 0.00403594970703125,\n -0.0009860992431640625,\n -0.027862548828125,\n 0.027587890625,\n 0.0006814002990722656,\n 0.047210693359375,\n -0.00902557373046875,\n -0.029144287109375,\n 0.00928497314453125,\n -0.004913330078125,\n 0.02557373046875,\n 0.00682830810546875,\n -0.01922607421875,\n -0.002452850341796875,\n -0.0266876220703125,\n 0.01000213623046875,\n -0.01494598388671875,\n 0.01554107666015625,\n 0.005794525146484375,\n -0.031829833984375,\n -0.0293731689453125,\n 0.01239013671875,\n 0.004276275634765625,\n 0.003032684326171875,\n 0.0150146484375,\n 0.00402069091796875,\n -0.01136016845703125,\n 0.006481170654296875,\n -0.01727294921875,\n -0.0292205810546875,\n -0.0304412841796875,\n -0.020965576171875,\n -0.00982666015625,\n -0.0179901123046875,\n -0.035186767578125,\n -0.0214996337890625,\n 0.03265380859375,\n 0.023345947265625,\n -0.025482177734375,\n -0.005985260009765625,\n 0.0023441314697265625,\n 0.006561279296875,\n 0.00400543212890625,\n 0.032958984375,\n -0.00566864013671875,\n 0.0191497802734375,\n 0.00011974573135375977,\n 0.01081085205078125,\n -0.0224609375,\n -0.0294342041015625,\n 0.03863525390625,\n -0.016326904296875,\n 0.03204345703125,\n 0.013916015625,\n 0.0017852783203125,\n 0.0191650390625,\n -0.0242767333984375,\n -0.00342559814453125,\n -0.0218963623046875,\n 0.019378662109375,\n -0.011199951171875,\n 0.009307861328125,\n -0.0198974609375,\n -0.00701904296875,\n -0.01364898681640625,\n -0.0290374755859375,\n -0.00849151611328125,\n 0.004833221435546875,\n -0.00485992431640625,\n -0.0012979507446289062,\n 0.0251007080078125,\n 0.0169525146484375,\n -0.00701904296875,\n -0.0051116943359375,\n -0.0169830322265625,\n 0.04351806640625,\n -0.041412353515625,\n 0.00689697265625,\n -0.020782470703125,\n -0.01416015625,\n -0.033599853515625,\n 0.0013570785522460938,\n -0.04241943359375,\n -0.027191162109375,\n -0.01019287109375,\n -0.0164337158203125,\n 0.0165557861328125,\n -0.01922607421875,\n 0.036651611328125,\n -0.0261993408203125,\n -0.0012950897216796875,\n 0.02484130859375,\n 0.03759765625,\n -0.00919342041015625,\n 0.0182952880859375,\n 0.0310821533203125,\n 0.0141754150390625,\n -0.015533447265625,\n -0.03033447265625,\n -0.021575927734375,\n -0.02410888671875,\n 0.00817108154296875,\n 0.0011730194091796875,\n -0.0014801025390625,\n -0.0079803466796875,\n 0.0198822021484375,\n -0.0037078857421875,\n 0.0474853515625,\n 0.035736083984375,\n 0.012115478515625,\n 0.033538818359375,\n -0.0001016855239868164,\n -0.02001953125,\n 0.01580810546875,\n 0.0146026611328125,\n -0.025970458984375,\n 0.004581451416015625,\n -0.005603790283203125,\n 0.04888916015625,\n 0.00835418701171875,\n 0.01727294921875,\n 0.0167694091796875,\n 0.021697998046875,\n 0.0158233642578125,\n -0.023773193359375,\n 0.0262908935546875,\n -0.013763427734375,\n -0.0123291015625,\n -0.01165008544921875,\n 0.0171966552734375,\n -0.06005859375,\n 0.00960540771484375,\n 0.0177154541015625,\n 0.008514404296875,\n 0.0167388916015625,\n -0.01123046875,\n -0.035003662109375,\n 0.0006346702575683594,\n 0.004352569580078125,\n -0.041778564453125,\n 0.025665283203125,\n -0.0308074951171875,\n -0.034271240234375,\n 0.0043487548828125,\n -0.00344085693359375,\n 0.006683349609375,\n 0.0484619140625,\n -0.01386260986328125,\n -0.0052032470703125,\n -0.005847930908203125,\n 0.012298583984375,\n 0.022247314453125,\n 0.0404052734375,\n 0.0068511962890625,\n 0.0027618408203125,\n 0.0280609130859375,\n -0.0023174285888671875,\n 0.02099609375,\n -0.008544921875,\n 0.0016622543334960938,\n -0.0293121337890625,\n 0.023223876953125,\n -0.024932861328125,\n -0.003261566162109375,\n -0.007568359375,\n -0.0011339187622070312,\n 0.02117919921875,\n -0.055999755859375,\n 0.01334381103515625,\n 0.0017671585083007812,\n 0.0071258544921875,\n 0.00896453857421875,\n 0.0186004638671875,\n 0.007701873779296875,\n 0.0129241943359375,\n 0.0203094482421875,\n 0.0408935546875,\n 0.0239410400390625,\n -0.0037479400634765625,\n 0.01381683349609375,\n -0.00028634071350097656,\n -0.01364898681640625,\n -0.03131103515625,\n -0.00901031494140625,\n -0.0050506591796875,\n -0.0028743743896484375,\n -0.017822265625,\n -0.035980224609375,\n 0.05218505859375,\n 0.020965576171875,\n 0.0021915435791015625,\n 0.010772705078125,\n -0.01262664794921875,\n -0.01100921630859375,\n -0.0082550048828125,\n 0.0079345703125,\n 0.001178741455078125,\n 0.012939453125,\n 0.00982666015625,\n 0.01983642578125,\n 0.0214385986328125,\n -0.00847625732421875,\n -0.0152130126953125,\n 0.005268096923828125,\n -0.00836944580078125,\n -0.0160369873046875,\n 0.0235748291015625,\n -0.047149658203125,\n 0.0374755859375,\n 0.01532745361328125,\n 0.007053375244140625,\n -0.046356201171875,\n 0.04119873046875,\n -0.01507568359375,\n 0.029144287109375,\n 0.013702392578125,\n -0.0211029052734375,\n 0.007167816162109375,\n -0.02618408203125,\n -0.03326416015625,\n -0.0197296142578125,\n -0.0008974075317382812,\n 0.0299224853515625,\n 0.00274658203125,\n 0.028076171875,\n -0.0142059326171875,\n 0.03033447265625,\n 0.0036678314208984375,\n -0.0145721435546875,\n -0.0167236328125,\n -0.00594329833984375,\n 0.0191497802734375,\n 0.0277252197265625,\n 0.003032684326171875,\n 0.045928955078125,\n -0.0062103271484375,\n 0.0010519027709960938,\n -0.0185089111328125,\n 0.015838623046875,\n 0.02471923828125,\n -0.01451873779296875,\n 0.020538330078125,\n 0.00485992431640625,\n 0.0504150390625,\n -0.031494140625,\n 0.0243988037109375,\n -0.0189361572265625,\n 0.0063934326171875,\n 0.0287628173828125,\n -0.0008535385131835938,\n -0.0187225341796875,\n -0.0005669593811035156,\n 0.0293426513671875,\n 0.006412506103515625,\n -0.00036835670471191406,\n -0.013702392578125,\n 0.003551483154296875,\n -0.011322021484375,\n 0.002254486083984375,\n -0.01708984375,\n -0.0132598876953125,\n -0.0029506683349609375,\n -0.0272369384765625,\n -0.0286102294921875,\n 0.029541015625,\n 0.007965087890625,\n -0.002719879150390625,\n -0.00835418701171875,\n -0.0009479522705078125,\n 0.0094146728515625,\n 0.019287109375,\n -0.004444122314453125,\n 0.01206207275390625,\n 0.029327392578125,\n 0.006359100341796875,\n -0.021697998046875,\n 0.0050048828125,\n -0.0084381103515625,\n 0.02423095703125,\n -0.004180908203125,\n -0.01302337646484375,\n -0.03497314453125,\n -0.01154327392578125,\n -0.016632080078125,\n -0.0011310577392578125,\n 0.024658203125,\n 0.031646728515625,\n -0.034912109375,\n -0.0015211105346679688,\n -0.033538818359375,\n 0.0194549560546875,\n -0.01015472412109375,\n 0.024810791015625,\n 0.0208282470703125,\n -0.00013458728790283203,\n -0.03466796875,\n 0.0038318634033203125,\n 0.0260772705078125,\n 0.003482818603515625,\n -0.0226898193359375,\n -0.0224609375,\n 0.032989501953125,\n 0.035919189453125,\n -0.022064208984375,\n 0.059234619140625,\n -0.0006074905395507812,\n -0.002620697021484375,\n 0.00531768798828125,\n 0.0274658203125,\n 0.005001068115234375,\n -0.022308349609375,\n -0.004520416259765625,\n 0.016326904296875,\n -0.015045166015625,\n 0.02288818359375,\n -0.0012025833129882812,\n 0.0246124267578125,\n -0.0018339157104492188,\n 0.0195159912109375,\n -0.01480865478515625,\n -0.0400390625,\n 0.0280609130859375,\n -0.0169830322265625,\n 0.02789306640625,\n -0.02679443359375,\n 0.012603759765625,\n 0.0200958251953125,\n -0.00168609619140625,\n 0.00876617431640625,\n -0.020355224609375,\n -0.044677734375,\n -0.0126495361328125,\n 0.006809234619140625,\n -0.012420654296875,\n 0.0279541015625,\n 0.0184173583984375,\n 0.05517578125,\n -0.0240020751953125,\n 0.042205810546875,\n 0.050628662109375,\n -0.034576416015625,\n 0.0014581680297851562,\n 0.033966064453125,\n 0.020782470703125,\n 0.003265380859375,\n -0.00975799560546875,\n 0.0268402099609375,\n -0.00138092041015625,\n -0.03369140625,\n 0.01226806640625,\n -0.01471710205078125,\n 0.00582122802734375,\n -0.00021183490753173828,\n 0.0147705078125,\n -0.00708770751953125,\n -0.0310821533203125,\n -0.01053619384765625,\n -0.01145172119140625,\n -0.0140228271484375,\n 0.04461669921875,\n -0.034820556640625,\n 0.011566162109375,\n -0.031982421875,\n 0.0033721923828125,\n 0.00960540771484375,\n 0.0015716552734375,\n 0.02728271484375,\n -0.01507568359375,\n 0.0304412841796875,\n 0.04083251953125,\n 0.0002872943878173828,\n -0.040252685546875,\n 0.007045745849609375,\n 0.005306243896484375,\n -0.006122589111328125,\n -0.00746917724609375,\n -0.0008921623229980469,\n 0.01529693603515625,\n -0.01029205322265625,\n 0.0161285400390625,\n -0.003849029541015625,\n -0.018280029296875,\n -0.00005936622619628906,\n 0.024993896484375,\n 0.01430511474609375,\n -0.050994873046875,\n 0.010223388671875,\n -0.0006861686706542969,\n -0.00809478759765625,\n -0.006824493408203125,\n 0.004261016845703125,\n -0.01491546630859375,\n -0.007266998291015625,\n -0.0171966552734375,\n 0.003070831298828125,\n -0.0399169921875,\n 0.0191802978515625,\n 0.0013837814331054688,\n -0.013916015625,\n 0.01302337646484375,\n 0.000019550323486328125,\n 0.0027923583984375,\n 0.0247039794921875,\n -0.009429931640625,\n -0.0028705596923828125,\n 0.00457000732421875,\n -0.0142822265625,\n -0.01148223876953125,\n 0.0052490234375,\n -0.0159149169921875,\n -0.0019388198852539062,\n -0.037506103515625,\n 0.040008544921875,\n -0.01375579833984375,\n -0.0196990966796875,\n -0.0284881591796875,\n -0.045379638671875,\n -0.0181732177734375,\n 0.0174560546875,\n 0.004459381103515625,\n 0.01271820068359375,\n 0.01934814453125,\n 0.009796142578125,\n -0.005062103271484375,\n 0.0093841552734375,\n -0.0208892822265625,\n 0.0121917724609375,\n -0.0177764892578125,\n -0.006221771240234375,\n 0.03680419921875,\n 0.01007080078125,\n -0.029144287109375,\n -0.031341552734375,\n -0.0173492431640625,\n 0.0003707408905029297,\n -0.0161590576171875,\n -0.00705718994140625,\n 0.006793975830078125,\n 0.0013456344604492188,\n -0.0104217529296875,\n 0.033233642578125,\n -0.032318115234375,\n 0.02642822265625,\n -0.0277252197265625,\n -0.01290130615234375,\n 0.037750244140625,\n -0.00661468505859375,\n -0.0141754150390625,\n -0.007701873779296875,\n -0.01427459716796875,\n 0.0267486572265625,\n 0.0296173095703125,\n 0.020538330078125,\n -0.00782012939453125,\n -0.00708770751953125,\n -0.00970458984375,\n 0.0216827392578125,\n -0.002849578857421875,\n -0.01108551025390625,\n 0.0034046173095703125,\n 0.03961181640625,\n -0.0018873214721679688,\n -0.04412841796875,\n 0.00012421607971191406,\n -0.034912109375,\n 0.0191650390625,\n 0.0181884765625,\n 0.0227508544921875,\n -0.005626678466796875,\n -0.01312255859375,\n 0.0309600830078125,\n -0.0082550048828125,\n -0.025146484375,\n 0.0254364013671875,\n 0.0180816650390625,\n 0.01177215576171875,\n -0.034423828125,\n -0.01605224609375,\n 0.0127105712890625,\n 0.004840850830078125,\n -0.0209503173828125,\n 0.0396728515625,\n 0.0267333984375,\n 0.037261962890625,\n -0.003734588623046875,\n 0.020263671875,\n -0.038604736328125,\n -0.0745849609375,\n 0.037811279296875,\n -0.006977081298828125,\n 0.01464080810546875,\n -0.00995635986328125,\n -0.0208587646484375,\n -0.0273590087890625,\n 0.007343292236328125,\n -0.0005130767822265625,\n -0.00595855712890625,\n -0.0200042724609375,\n -0.025604248046875,\n -0.003459930419921875,\n 0.0141754150390625,\n 0.002593994140625,\n 0.0340576171875,\n 0.015167236328125,\n 0.002628326416015625,\n -0.0027828216552734375,\n -0.0104827880859375,\n 0.033843994140625,\n 0.028045654296875,\n 0.019622802734375,\n 0.0241241455078125,\n 0.01245880126953125,\n 0.0047607421875,\n -0.01715087890625,\n 0.02154541015625,\n 0.01226806640625,\n -0.0106353759765625,\n 0.056640625,\n 0.0219879150390625,\n 0.0202789306640625,\n -0.00531768798828125,\n -0.0143585205078125,\n -0.00009465217590332031,\n -0.01409149169921875,\n 0.013519287109375,\n 0.007480621337890625,\n 0.00637054443359375,\n -0.04168701171875,\n 0.01230621337890625,\n 0.037017822265625,\n -0.030548095703125,\n -0.01399993896484375,\n -0.006072998046875,\n -0.0183258056640625,\n -0.0284881591796875,\n -0.02154541015625,\n 0.004627227783203125,\n -0.01149749755859375,\n -0.019775390625,\n 0.037261962890625,\n -0.01158905029296875,\n -0.004848480224609375,\n -0.0293426513671875,\n -0.004528045654296875,\n -0.0015659332275390625,\n -0.020111083984375,\n 0.020111083984375,\n -0.021148681640625,\n -0.036773681640625,\n 0.004878997802734375,\n 0.0130767822265625,\n -0.002410888671875,\n -0.015960693359375,\n 0.037353515625,\n -0.008148193359375,\n -0.06597900390625,\n 0.03448486328125,\n -0.006744384765625,\n 0.01416015625,\n -0.0267333984375,\n 0.02886962890625,\n 0.0133209228515625,\n -0.0369873046875,\n -0.039459228515625,\n 0.028717041015625,\n 0.0205230712890625,\n 0.038330078125,\n -0.01535797119140625,\n 0.027587890625,\n 0.023223876953125,\n 0.003604888916015625,\n -0.0195465087890625,\n 0.01029205322265625,\n 0.004791259765625,\n 0.017059326171875,\n -0.0101165771484375,\n -0.035125732421875,\n 0.03857421875,\n 0.00849151611328125,\n 0.048675537109375,\n 0.0141754150390625,\n 0.0080413818359375,\n 0.0063018798828125,\n 0.006786346435546875,\n -0.0256500244140625,\n 0.0132904052734375,\n -0.007110595703125,\n -0.023406982421875,\n -0.049224853515625,\n -0.029327392578125,\n -0.0132598876953125,\n -0.00830078125,\n -0.00479888916015625,\n 0.0306396484375,\n -0.024383544921875,\n 0.0174560546875,\n -0.0092315673828125,\n -0.0196990966796875,\n -0.036376953125,\n -0.0027618408203125,\n 0.0029201507568359375,\n 0.001956939697265625,\n -0.0013761520385742188,\n 0.044708251953125,\n 0.00916290283203125,\n 0.0122833251953125,\n 0.007659912109375,\n 0.01381683349609375,\n -0.0430908203125,\n 0.0170135498046875,\n 0.028717041015625,\n -0.00998687744140625,\n 0.00720977783203125,\n 0.027740478515625,\n -0.01158905029296875,\n 0.0310821533203125,\n -0.01019287109375,\n 0.0019464492797851562,\n -0.033050537109375,\n 0.01126861572265625,\n 0.0130462646484375,\n 0.01861572265625,\n 0.007335662841796875,\n 0.027801513671875,\n 0.00594329833984375,\n 0.0122222900390625,\n 0.007232666015625,\n -0.033905029296875,\n 0.0274658203125,\n 0.001674652099609375,\n -0.027191162109375,\n -0.001514434814453125,\n -0.04278564453125,\n -0.0177001953125,\n 0.048187255859375,\n -0.03070068359375,\n -0.02191162109375,\n 0.031585693359375,\n 0.028900146484375,\n -0.00907135009765625,\n -0.020477294921875,\n 0.0037555694580078125,\n -0.0008893013000488281,\n -0.01071929931640625,\n 0.0199127197265625,\n -0.044403076171875,\n -0.049102783203125,\n -0.0143585205078125,\n -0.019073486328125,\n -0.0227508544921875,\n 0.0302734375,\n -0.00743865966796875,\n 0.005764007568359375,\n 0.00872802734375,\n 0.013519287109375,\n -0.0166168212890625,\n 0.0008749961853027344,\n -0.028076171875,\n 0.031280517578125,\n -0.002857208251953125,\n -0.006481170654296875,\n -0.0023212432861328125,\n -0.01312255859375,\n -0.02203369140625,\n -0.005268096923828125,\n -0.0103607177734375,\n -0.051422119140625,\n -0.0018215179443359375,\n 0.03460693359375,\n 0.0140533447265625,\n 0.00910186767578125,\n -0.011444091796875,\n -0.0007572174072265625,\n 0.0173492431640625,\n 0.01410675048828125,\n 0.016082763671875,\n 0.0048675537109375,\n -0.0283966064453125,\n 0.019622802734375,\n -0.01019287109375,\n 0.019805908203125,\n -0.0606689453125,\n -0.00238037109375,\n -0.0231170654296875,\n -0.00975799560546875,\n 0.010589599609375,\n 0.0298309326171875,\n -0.02703857421875,\n 0.03955078125,\n 0.0193634033203125,\n -0.021453857421875,\n 0.00897979736328125,\n -0.00994110107421875,\n 0.029937744140625,\n 0.00797271728515625,\n -0.08343505859375,\n -0.019073486328125,\n 0.0016622543334960938,\n -0.0122222900390625,\n -0.028106689453125,\n -0.0204925537109375,\n 0.01290130615234375,\n 0.047088623046875,\n 0.035858154296875,\n -0.006786346435546875,\n 0.01134490966796875,\n 0.02703857421875,\n -0.007747650146484375,\n 0.00020372867584228516,\n -0.0120849609375,\n -0.00899505615234375,\n 0.0016374588012695312,\n 0.01509857177734375,\n -0.01056671142578125,\n 0.0178375244140625,\n -0.03466796875,\n 0.032318115234375,\n -0.013763427734375,\n -0.0212554931640625,\n 0.00748443603515625,\n 0.0274658203125,\n -0.0164031982421875,\n -0.0208282470703125,\n -0.026458740234375,\n 0.0279541015625,\n -0.01727294921875,\n -0.0235137939453125,\n 0.009552001953125,\n 0.00211334228515625,\n -0.00917816162109375,\n 0.033050537109375,\n 0.005901336669921875,\n 0.0107574462890625,\n -0.01245880126953125,\n -0.00200653076171875,\n 0.0179901123046875,\n -0.005886077880859375,\n -0.01554107666015625,\n -0.034027099609375,\n 0.0009660720825195312,\n 0.0111236572265625,\n -0.0152587890625,\n 0.020721435546875,\n 0.028961181640625,\n -0.0043487548828125,\n -0.0024852752685546875,\n -0.0218505859375,\n 0.0007123947143554688,\n -0.0426025390625,\n -0.0006699562072753906,\n -0.00372314453125,\n -0.0248870849609375,\n -0.03167724609375,\n 0.0140533447265625,\n -0.01213836669921875,\n -0.00481414794921875,\n -0.0262451171875,\n -0.000385284423828125,\n -0.006072998046875,\n -0.004093170166015625,\n -0.06793212890625,\n -0.05712890625,\n -0.021209716796875,\n -0.0140533447265625,\n -0.03466796875,\n -0.0144500732421875,\n 0.01158905029296875,\n -0.0310211181640625,\n 0.03692626953125,\n -0.02899169921875,\n -0.00197601318359375,\n -0.0003044605255126953,\n 0.0013093948364257812,\n 0.0297088623046875,\n -0.00620269775390625,\n 0.005146026611328125,\n -0.01111602783203125,\n -0.00632476806640625,\n -0.02740478515625,\n -0.03729248046875,\n 0.0017337799072265625,\n 0.00995635986328125,\n -0.026458740234375,\n -0.006153106689453125,\n -0.00337982177734375,\n 0.018096923828125,\n 0.0384521484375,\n -0.00411224365234375,\n -0.033905029296875,\n -0.00334930419921875,\n -0.011505126953125,\n -0.020721435546875,\n -0.005702972412109375,\n -0.0042266845703125,\n 0.0078277587890625,\n 0.01265716552734375,\n -0.046356201171875,\n -0.0162200927734375,\n 0.007526397705078125,\n -0.015960693359375,\n -0.035552978515625,\n -0.003742218017578125,\n -0.005130767822265625,\n 0.020660400390625,\n -0.00572967529296875,\n -0.0260162353515625,\n 0.01235198974609375,\n 0.00004369020462036133,\n -0.0025691986083984375,\n 0.004543304443359375,\n 0.027587890625,\n -0.02783203125,\n -0.00531005859375,\n -0.007358551025390625,\n -0.0205841064453125,\n -0.044677734375,\n -0.00797271728515625,\n 0.006359100341796875,\n -0.0227508544921875,\n 0.0068359375,\n -0.03009033203125,\n -0.026763916015625,\n 0.01629638671875,\n -0.04046630859375,\n 0.012115478515625,\n -0.026031494140625,\n -0.01427459716796875,\n 0.023223876953125,\n -0.00116729736328125,\n 0.01038360595703125,\n -0.01216888427734375,\n 0.022064208984375,\n -0.023712158203125,\n 0.0279541015625,\n -0.005550384521484375,\n -0.0253143310546875,\n 0.00556182861328125,\n -0.029205322265625,\n -0.013214111328125,\n -0.01055145263671875,\n -0.01328277587890625,\n 0.0128021240234375,\n 0.0364990234375,\n 0.0180511474609375,\n -0.01708984375,\n 0.03228759765625,\n 0.0110626220703125,\n 0.00887298583984375,\n 0.01409149169921875,\n 0.005474090576171875,\n -0.00045371055603027344,\n -0.01168060302734375,\n 0.0130767822265625,\n -0.016632080078125,\n 0.0020599365234375,\n 0.022125244140625,\n -0.0035610198974609375,\n -0.0015239715576171875,\n 0.015655517578125,\n -0.037933349609375,\n 0.0235748291015625,\n -0.00960540771484375,\n -0.004741668701171875,\n 0.016265869140625,\n -0.005481719970703125,\n 0.01361846923828125,\n -0.02520751953125,\n -0.004581451416015625,\n 0.033111572265625,\n 0.0127410888671875,\n 0.02789306640625,\n -0.020172119140625,\n -0.013092041015625,\n 0.047454833984375,\n 0.0007686614990234375,\n -0.007740020751953125,\n 0.01328277587890625,\n 0.00412750244140625,\n -0.042877197265625,\n 0.01065826416015625,\n -0.00838470458984375,\n -0.020965576171875,\n 0.0203094482421875,\n 0.0104827880859375,\n 0.04010009765625,\n 0.021209716796875,\n -0.004917144775390625,\n -0.0108795166015625,\n 0.01244354248046875,\n 0.00846099853515625,\n 0.0088958740234375,\n -0.0037078857421875,\n -0.002422332763671875,\n -0.0440673828125,\n 0.0185089111328125,\n 0.004184722900390625,\n 0.048004150390625,\n -0.011993408203125,\n 0.0094451904296875,\n 0.005825042724609375,\n -0.0191497802734375,\n -0.0226287841796875,\n -0.0003876686096191406,\n -0.0022983551025390625,\n -0.00220489501953125,\n 0.00734710693359375,\n 0.0026340484619140625,\n 0.0244903564453125,\n -0.0172119140625,\n 0.0268096923828125,\n -0.034759521484375,\n 0.030029296875,\n 0.021453857421875,\n -0.020965576171875,\n 0.0146942138671875,\n -0.034210205078125,\n 0.01535797119140625,\n 0.034423828125,\n -0.0072479248046875,\n 0.00457000732421875,\n 0.04669189453125,\n -0.0341796875,\n 0.002040863037109375,\n -0.0019464492797851562,\n -0.0034046173095703125,\n 0.0020236968994140625,\n 0.0074310302734375,\n -0.01708984375,\n -0.037017822265625,\n -0.0146026611328125,\n -0.03546142578125,\n -0.0185089111328125,\n 0.01056671142578125,\n -0.040313720703125,\n -0.050506591796875,\n -0.00820159912109375,\n -0.0546875,\n -0.023681640625,\n -0.02484130859375,\n -0.0087127685546875,\n 0.03460693359375,\n 0.00009971857070922852,\n 0.0242919921875,\n -0.020416259765625,\n -0.0251007080078125,\n 0.0276641845703125,\n -0.008026123046875,\n -0.009521484375,\n 0.01690673828125,\n 0.000004410743713378906,\n -0.014251708984375,\n -0.002689361572265625,\n 0.011749267578125,\n 0.0197296142578125,\n -0.0276947021484375,\n 0.00514984130859375,\n 0.033905029296875,\n -0.0092926025390625,\n -0.0033512115478515625,\n -0.034088134765625,\n 0.023834228515625,\n -0.0019474029541015625,\n 0.00121307373046875,\n 0.019927978515625,\n -0.0287628173828125,\n -0.033538818359375,\n 0.006519317626953125,\n -0.0119476318359375,\n 0.0069427490234375,\n 0.050933837890625,\n -0.00904083251953125,\n -0.00496673583984375,\n 0.028961181640625,\n 0.00037407875061035156,\n 0.042266845703125,\n 0.0181732177734375,\n 0.0036163330078125,\n 0.01629638671875,\n 0.0158233642578125,\n 0.007030487060546875,\n -0.004718780517578125,\n 0.0017242431640625,\n -0.038909912109375,\n 0.0083465576171875,\n 0.00966644287109375,\n -0.004581451416015625,\n -0.0054931640625,\n 0.00978851318359375,\n 0.01027679443359375,\n -0.013702392578125,\n -0.0123748779296875,\n 0.0063934326171875,\n 0.027191162109375,\n -0.0309295654296875,\n -0.0217437744140625,\n -0.0249176025390625,\n 0.01073455810546875,\n -0.01218414306640625,\n -0.0238800048828125,\n 0.045501708984375,\n -0.028961181640625,\n 0.006290435791015625,\n 0.0095367431640625,\n -0.0235595703125,\n 0.005435943603515625,\n -0.036346435546875,\n -0.0163726806640625,\n -0.019012451171875,\n -0.0400390625,\n 0.0087127685546875,\n 0.01168060302734375,\n -0.019927978515625,\n -0.019317626953125,\n -0.0009331703186035156,\n -0.0004107952117919922,\n -0.0026073455810546875,\n 0.00342559814453125,\n 0.04571533203125,\n 0.007678985595703125,\n -0.0190887451171875,\n -0.01219940185546875,\n -0.020172119140625,\n -0.040771484375,\n 0.0132904052734375,\n 0.0028133392333984375,\n 0.006771087646484375,\n 0.004131317138671875,\n -0.0440673828125,\n 0.031585693359375,\n -0.0219879150390625,\n -0.026214599609375,\n 0.0290374755859375,\n -0.02520751953125,\n 0.0197601318359375,\n 0.02978515625,\n -0.015167236328125,\n -0.01180267333984375,\n -0.0316162109375,\n 0.001735687255859375,\n 0.01458740234375,\n -0.0009775161743164062,\n 0.0266265869140625,\n -0.0122222900390625,\n 0.0231170654296875,\n 0.012420654296875,\n 0.0240631103515625,\n -0.035980224609375,\n 0.03924560546875,\n -0.012420654296875,\n -0.0141754150390625,\n 0.02978515625,\n 0.01214599609375,\n 0.0256500244140625,\n -0.01507568359375,\n 0.0213775634765625,\n -0.00876617431640625,\n 0.0202789306640625,\n 0.0231170654296875,\n 0.0080413818359375,\n -0.0194854736328125,\n 0.00756072998046875,\n 0.0164031982421875,\n -0.0306549072265625,\n 0.0017375946044921875,\n 0.016357421875,\n -0.0054779052734375,\n 0.0240020751953125,\n -0.0023250579833984375,\n -0.01039886474609375,\n -0.0044097900390625,\n 0.0051727294921875,\n 0.0347900390625,\n 0.0203399658203125,\n -0.002277374267578125,\n -0.02691650390625,\n 0.01123046875,\n -0.0230712890625,\n 0.01221466064453125,\n -0.031951904296875,\n -0.0132904052734375,\n 0.0015573501586914062,\n 0.0081329345703125,\n -0.016326904296875,\n 0.00421142578125,\n 0.016845703125,\n 0.030548095703125,\n -0.0011539459228515625,\n -0.0196685791015625,\n -0.006153106689453125,\n 0.01003265380859375,\n -0.0030612945556640625,\n 0.03466796875,\n -0.01178741455078125,\n 0.0218658447265625,\n 0.01183319091796875,\n -0.015838623046875,\n -0.00782012939453125,\n -0.003528594970703125,\n -0.0225677490234375,\n 0.0199432373046875,\n -0.007427215576171875,\n 0.025604248046875,\n 0.0213470458984375,\n 0.0078582763671875,\n 0.01039886474609375,\n 0.021240234375,\n -0.012542724609375,\n 0.01611328125,\n 0.0313720703125,\n 0.00435638427734375,\n -0.0023021697998046875,\n -0.02069091796875,\n -0.0005769729614257812,\n 0.031585693359375,\n -0.005084991455078125,\n 0.0260162353515625,\n -0.012298583984375,\n -0.00618743896484375,\n 0.0129241943359375,\n -0.0419921875,\n -0.032318115234375,\n 0.01336669921875,\n 0.0343017578125,\n 0.037506103515625,\n -0.00975799560546875,\n 0.01068878173828125,\n -0.006450653076171875,\n -0.0036182403564453125,\n 0.0528564453125,\n 0.01364898681640625,\n 0.054290771484375,\n -0.041229248046875,\n 0.035980224609375,\n -0.01131439208984375,\n 0.022918701171875,\n 0.0023136138916015625,\n 0.041046142578125,\n -0.03228759765625,\n 0.033447265625,\n 0.0098114013671875,\n 0.01471710205078125,\n -0.01451873779296875,\n -0.003505706787109375,\n -0.0007944107055664062,\n 0.00888824462890625,\n -0.014434814453125,\n -0.01560211181640625,\n 0.039581298828125,\n -0.036712646484375,\n -0.01690673828125,\n 0.015777587890625,\n -0.00433349609375,\n -0.0199432373046875,\n -0.00594329833984375,\n -0.00966644287109375,\n -0.006195068359375,\n 0.0047454833984375,\n -0.0246124267578125,\n 0.01549530029296875,\n -0.0179290771484375,\n 0.053466796875,\n 0.051055908203125,\n 0.0242919921875,\n 0.0237884521484375,\n 0.01090240478515625,\n -0.02557373046875,\n -0.00540924072265625,\n 0.0299835205078125,\n 0.032684326171875,\n 0.0180511474609375,\n 0.0338134765625,\n 0.000301361083984375,\n -0.020538330078125,\n 0.0290374755859375,\n 0.01155853271484375,\n 0.05328369140625,\n -0.01016998291015625,\n 0.0048675537109375,\n 0.0265655517578125,\n 0.006988525390625,\n -0.02008056640625,\n -0.02459716796875,\n -0.04034423828125,\n -0.0001156926155090332,\n -0.017120361328125,\n -0.0312042236328125,\n -0.08050537109375,\n 0.009246826171875,\n -0.01177215576171875,\n 0.01433563232421875,\n -0.0005059242248535156,\n -0.01898193359375,\n 0.0146026611328125,\n 0.00020968914031982422,\n -0.0282440185546875,\n -0.020721435546875,\n -0.0078887939453125,\n 0.00934600830078125,\n 0.01219940185546875,\n 0.02130126953125,\n 0.036163330078125,\n -0.00826263427734375,\n 0.0126495361328125,\n 0.0166473388671875,\n 0.00372314453125,\n -0.0018815994262695312,\n 0.038543701171875,\n 0.0010766983032226562,\n -0.0098114013671875,\n -0.0132598876953125,\n -0.0217132568359375,\n -0.0137939453125,\n -0.00937652587890625,\n 0.028900146484375,\n -0.0271453857421875,\n -0.007259368896484375,\n -0.01959228515625,\n -0.00433349609375,\n -0.0205535888671875,\n -0.0183258056640625,\n -0.0127105712890625,\n 0.006977081298828125,\n 0.0018367767333984375,\n -0.0030155181884765625,\n 0.00959014892578125,\n -0.00818634033203125,\n 0.01108551025390625,\n -0.0304107666015625,\n -0.0114593505859375,\n 0.0219879150390625,\n -0.00087738037109375,\n -0.0239410400390625,\n 0.00443267822265625,\n -0.0012035369873046875,\n 0.034759521484375,\n 0.01416015625,\n -0.037567138671875,\n 0.031280517578125,\n -0.006862640380859375,\n -0.02508544921875,\n -0.0161895751953125,\n 0.02728271484375,\n -0.01139068603515625,\n 0.00475311279296875,\n 0.000667572021484375,\n 0.0225677490234375,\n -0.00963592529296875,\n 0.025482177734375,\n -0.018463134765625,\n -0.039581298828125,\n -0.0408935546875,\n -0.0138702392578125,\n 0.0260772705078125,\n 0.020965576171875,\n 0.030792236328125,\n 0.028167724609375,\n -0.0185546875,\n 0.004425048828125,\n -0.028594970703125,\n 0.00897979736328125,\n -0.01424407958984375,\n -0.0096893310546875,\n 0.0020999908447265625,\n -0.01512908935546875,\n 0.026458740234375,\n 0.005672454833984375,\n 0.0175628662109375,\n 0.005825042724609375,\n -0.0183563232421875,\n 0.0216827392578125,\n -0.029754638671875,\n -0.0222015380859375,\n -0.003185272216796875,\n 0.00878143310546875,\n -0.0152435302734375,\n 0.006740570068359375,\n 0.006793975830078125,\n -0.003658294677734375,\n 0.042510986328125,\n -0.02691650390625,\n 0.0237884521484375,\n -0.0021228790283203125,\n -0.004825592041015625,\n -0.017059326171875,\n 0.054901123046875,\n 0.0268096923828125,\n 0.04888916015625,\n 0.013641357421875,\n 0.02520751953125,\n 0.016021728515625,\n 0.00025916099548339844,\n 0.027130126953125,\n 0.034912109375,\n 0.0085296630859375,\n -0.0081329345703125,\n -0.00397491455078125,\n -0.01399993896484375,\n -0.0082244873046875,\n 0.01397705078125,\n -0.005687713623046875,\n -0.01282501220703125,\n 0.0172882080078125,\n 0.025146484375,\n -0.0218963623046875,\n -0.0019073486328125,\n 0.01097869873046875,\n 0.0016794204711914062,\n -0.002399444580078125,\n -0.0085296630859375,\n 0.0281829833984375,\n 0.0252532958984375,\n -0.0009751319885253906,\n -0.007537841796875,\n -0.020172119140625,\n -0.0159912109375,\n -0.0180511474609375,\n 0.0071258544921875,\n 0.004245758056640625,\n -0.0142822265625,\n -0.019012451171875,\n 0.0181884765625,\n -0.032958984375,\n -0.0211334228515625,\n -0.0255279541015625,\n -0.0191802978515625,\n -0.045989990234375,\n -0.00395965576171875,\n -0.004886627197265625,\n -0.0214996337890625,\n 0.02154541015625,\n 0.007537841796875,\n 0.055908203125,\n -0.01203155517578125,\n -0.0157623291015625,\n 0.00783538818359375,\n 0.012908935546875,\n -0.0175628662109375,\n -0.03204345703125,\n -0.01296234130859375,\n 0.03265380859375,\n 0.01934814453125,\n -0.00917816162109375,\n -0.0208892822265625,\n -0.0281219482421875,\n -0.0008893013000488281,\n -0.0077667236328125,\n -0.01224517822265625,\n -0.00033354759216308594,\n -0.006439208984375,\n -0.01094818115234375,\n -0.036956787109375,\n -0.023651123046875,\n -0.0230560302734375,\n 0.0165557861328125,\n 0.01922607421875,\n -0.008819580078125,\n -0.0121002197265625,\n 0.027313232421875,\n 0.007122039794921875,\n -0.033843994140625,\n -0.01560211181640625,\n -0.026397705078125,\n 0.0162200927734375,\n 0.01751708984375,\n 0.0229644775390625,\n -0.017822265625,\n 0.0267486572265625,\n 0.044403076171875,\n -0.018798828125,\n 0.03265380859375,\n 0.02349853515625,\n -0.041351318359375,\n -0.05206298828125,\n 0.01296234130859375,\n -0.0015277862548828125,\n -0.0010213851928710938,\n -0.0303802490234375,\n 0.0015411376953125,\n 0.00531005859375,\n 0.02239990234375,\n -0.007843017578125,\n -0.01262664794921875,\n 0.02899169921875,\n 0.036651611328125\n ],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/llama-3.2-nv-embedqa-1b-v2\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"bd:T46c,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/llama-3.2-nv-embedqa-1b-v2:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X \"POST\" \\\n \"http://localhost:8000/v1/embeddings\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\"Hello world\"],\n \"model\": \"nvidia/llama-3.2-nv-embedqa-1b-v2\",\n \"input_type\": \"query\"\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/overview.html).be:Tec6,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/support-matrix.html#llama-3-2-nv-embedqa-1b-v2))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench -u root\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod -R a+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/llama-3.2-nv-embedqa-1b-v2:1.4.0-rtx\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X \"POST\" \\\n \"http://localhost:8000/v1/embeddings\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\"Hello world\"],\n \"model\": \"nvidia/llama-3.2-nv-embedqa-1b-v2\",\n \"input_type\": \"query\"\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/getting-started.html)\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;"])</script><script>self.__next_f.push([1,"bf:T1309,"])</script><script>self.__next_f.push([1,"# NV-CLIP (Commercial Foundation Model)\n\n## Model Overview\nNV-CLIP is a multimodal embeddings model for image and text. Trained on 700M proprietary images, NV-CLIP is the NVIDIA commercial version of OpenAI’s CLIP (Contrastive Language-Image Pre-Training) model. NV-CLIP can be applied to various areas such as multimodal search, zero-shot image classification, and downstream computer vision tasks such as object detection and more.\n\n## Getting Started with NV-CLIP NIM microservice\n\nDeploying and integrating NV-CLIP NIM microservice is straightforward and based on industry standard APIs. See the NV-CLIP NIM microservice documentation to get started.\n\n## Applications\n\n- Multimodal search: Enable accurate image and text search to quickly search database of images and videos.\n\n- Zero-shot and few-shot inference: Classify images without re-training or fine-tuning.\n\n- Downstream vision tasks: Use the embeddings to enable downstream complex vision AI tasks such as segmentation, detection, VLMs and more.\n\n## References:\n\n- Radford, Alec, et al. \"Learning transferable visual models from natural language supervision.\" International conference on machine learning. PMLR, 2021.\n\n## Model Architecture: \n**Architecture Type:** Transformer-Based \u003cbr\u003e\n\nNV-CLIP as a backbone can be used towards various downstream tasks such as classification, detection, segmentation and text based image retrieval.\n\n## Input:\n**Input Type(s):** Images, Texts \u003cbr\u003e\n**Input Format(s):** List of Red, Green, Blue (RGB) Images or Strings \u003cbr\u003e\n**Other Properties Related to Input:** \u003cbr\u003e\n\nChannel Ordering of the Input: NCHW, where N = Batch Size, C = number of channels (3), H = Height of images (224), W = Width of the images (224)\n \n\n## Output:\n**Output Type(s):** Float tensor \u003cbr\u003e\n**Output Format:** 3D Tensor \u003cbr\u003e\n**Other Properties Related to Output:** \u003cbr\u003e \nThe output of this model is an embedding of an input image or text of size 1024 for ViT-H variant.\n\n\n**Supported Operating System(s):** \u003cbr\u003e\n* Linux \u003cbr\u003e\n\n## Model Version(s): \n- **nv_clip_224_vit_h_trainable_v1.0** - NV-CLIP ViT-H with 224 resolution is foundation model and is trainable.\n\n## Using this Model \u003ca class=\"anchor\" name=\"how_to_use_this_model\"\u003e\u003c/a\u003e\n\nThese models need to be used with NVIDIA hardware and software. These models can only be used with NV-CLIP NIM microservice.\n\nThe primary use case for these models is getting feature embeddings from images and text. These embeddings can then be used for curation, clustering, zero-shot or few-shot downstream tasks such as classification. These embeddings can also be used towards text and image-based image\n\n## Training Dataset:\n\n**Data Collection Method by dataset:** \u003cbr\u003e\n* Automated \u003cbr\u003e\n\n**Labeling Method by dataset:** \u003cbr\u003e\n* Automated \u003cbr\u003e\n\n**Properties:** \u003cbr\u003e\n\n| Dataset | No. of Images |\n|--|--|\n|NV Internal Data| 700M | \n\n\n## Evaluation Dataset:\n\n**Link:** [https://www.image-net.org/](https://www.image-net.org/)\n\n**Data Collection Method by dataset:** \u003cbr\u003e\n* Unknown \u003cbr\u003e\n\n**Labeling Method by dataset:** \u003cbr\u003e\n* Unknown \u003cbr\u003e\n\n**Properties:** \u003cbr\u003e\n50,000 validation images from [ImageNet dataset](https://www.image-net.org/download.php) \n\n#### Methodology and KPI\n\nThe performance of zero shot accuracy of NV-CLIP on ImageNet validation dataset.\n\n| model | top-1 Accuracy |\n| ----------------------- | -------------- |\n| ViT-H-224 | 0.7786 |\n\n## Ethical Considerations:\n### Bias, Safety \u0026 Security, and Privacy\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards [here](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/tao/models/nvclip_vit/bias). Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n### Special Training Data Considerations\n\nThe model was trained on publicly available data, which may contain toxic language and societal biases. Therefore, the model may amplify those biases, such as associating certain genders with specific social stereotypes.\n\n#### Governing Terms \nThe NIM container is governed by the NVIDIA AI Enterprise Software License Agreement | NVIDIA; and the use of this model is governed by the ai-foundation-models-community-license.pdf (nvidia.com).\n\n**You are responsible for ensuring that your use of NVIDIA AI Foundation Models complies with all applicable laws.**"])</script><script>self.__next_f.push([1,"c0:T48b,| Field |Response |\n|:-------------------|:---------|\n| Intended Application \u0026 Domain:| Generating image embedding that is aligned with text for zero-shot classification. |\n| Model Type: | Embedding Generation |\n| Intended User: | This model is intended for developers building search engines, classification, detection/ segmentation models. |\n| Output: | Embedding (An array of float numbers, providing a dense vector representation for the input text or image) |\n| Describe how the model works: | NV-CLIP consists of a vision and text encoder. The model maps both images and text into a unified embedding space. |\n| Verified to have met prescribed NVIDIA quality standards:| Yes|\n| Performance Metrics: | ImageNet zero-shot accuracy |\n| Potential Known Risks: | Unknown |\n| Licensing: | [NVIDIA AI Foundation Models Community License](https://docs.nvidia.com/ai-foundation-models-community-license.pdf) |\n| Technical Limitations: | It may not perform well in classifying images of aircraft, geolocation features, and faces. |c1:T4e5,| Field | Response |\n| -- | -- |\n| Generatable or reverse engineerable personally-identifiable information (PII)? | None |\n| Protected classes used to create this model? | Not Applicable (No PII) |\n| Was consent obtained for any PII used? | Not Applicable (No PII) |\n| How often is dataset reviewed? | \tBefore Release |\n| Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n| If PII collected for the development of the model, was it collected directly by NVIDIA? |Not Applicable |\n| If PII collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects?\t| Not Applicable |\n| If PII collected for the development of this AI model, was it minimized to only what was required? | Not Applicable |\n| Is there provenance for all datasets used in training? | Yes |\n| Do"])</script><script>self.__next_f.push([1,"es data labeling (annotation, metadata) comply with privacy laws? | Yes |\n| Is data compliant with data subject requests for data correction or removal, if such a request was made? | Yes |\n| Applicable NVIDIA Privacy Policy\t| [https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) |c2:T74a5,"])</script><script>self.__next_f.push([1,"curl -X POST https://integrate.api.nvidia.com/v1/embeddings \\\n -H \"Content-Type: application/json\" \\\n -H \"Authorization: Bearer $NVIDIA_API_KEY\" \\\n -d '{\n \"input\": [\"Image of a dog\",\n \"data:image/jpeg;base64,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\"],\n \"model\": \"nvidia/nvclip\",\n \"encoding_format\": \"float\"\n }'\n"])</script><script>self.__next_f.push([1,"c3:T74c0,"])</script><script>self.__next_f.push([1,"from openai import OpenAI\n\nclient = OpenAI(\n api_key=\"$NVIDIA_API_KEY\",\n base_url=\"https://integrate.api.nvidia.com/v1\"\n)\n\nresponse = client.embeddings.create(\n input=[ \"Image of a dog\",\n \"data:image/jpeg;base64,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\"],\n model=\"nvidia/nvclip\",\n encoding_format=\"float\"\n)\n\nprint(response.data[0].embedding)\n"])</script><script>self.__next_f.push([1,"c4:T7559,"])</script><script>self.__next_f.push([1,"import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1/',\n})\n\nasync function main() {\n const response = await openai.embeddings.create({\n input: [\"The quick brown fox jumped over the lazy dog\",\n \"data:image/jpeg;base64,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\"\n ],\n model: \"nvidia/nvclip\",\n encoding_format: \"float\"\n })\n\n const embedding = response.data[0].embedding;\n process.stdout.write(`${embedding}\\n`);\n}\n\nmain();\n"])</script><script>self.__next_f.push([1,"c5:Tab57,"])</script><script>self.__next_f.push([1,"{\n \"object\":\"list\",\n \"data\":[{\"index\":0,\"embedding\":[-0.004141246434301138,-0.01689707487821579,0.012426131404936314,-0.016726966947317123,-0.008678196929395199,-0.009295804426074028,0.009726164862513542,-0.022503215819597244,-0.017610320821404457,0.0003413025406189263,-0.03806029260158539,-0.0029619780834764242,0.03500921279191971,0.00837829802185297,0.0148561280220747,-0.0212395116686821,0.017078636214137077,-0.00459861196577549,-0.021547796204686165,-0.05555511638522148,-0.025865912437438965,-0.0058914124965667725,0.01959407702088356,0.03975767642259598,-0.007950921542942524,0.0019562202505767345,-0.04658181592822075,0.006377906538546085,0.015079125761985779,-0.031968407332897186,-0.03715363144874573,-0.0035252058878540993,0.012734952382743359,0.00885448046028614,-0.033137913793325424,0.004123990423977375,0.03223686292767525,-0.007955869659781456,-0.05179435387253761,0.02788241021335125,0.03838478401303291,0.004643972031772137,-0.000478812464280054,-0.02204490825533867,-0.01571882702410221,0.029129542410373688,-0.0157241839915514,-0.0316602922976017,0.033637937158346176,-0.029353737831115723,-0.012587918899953365,0.025821227580308914,-0.0188415739685297,0.0030159130692481995,0.009031707420945168,-0.01835968717932701,-0.004277291242033243,-0.015501229092478752,0.008056040853261948,0.02946413680911064,0.012887547723948956,0.03265497088432312,0.009581572376191616,-0.1564253270626068,-0.01405376847833395,0.010441438294947147,-0.02548121102154255,0.017590057104825974,0.028301002457737923,0.0026452515739947557,0.013897688128054142,-0.014540997333824635,-0.004527798388153315,0.03835851326584816,-0.03831495717167854,0.019475141540169716,0.010080724954605103,0.02765466645359993,0.00603770837187767,-0.000400348479161039,-0.036044176667928696,-0.012854951433837414,0.03269652649760246,0.010578403249382973,-0.0032332660630345345,-0.0013423014897853136,0.01841983199119568,0.009791318327188492,0.014436286874115467,0.0052145193330943584,-0.03309307619929314,-0.00016763672465458512,-0.0008141322759911418,0.01871390827000141,0.025392455980181694,0.0025894346181303263,0.04925240948796272,0.03293110802769661,-0.008387845940887928,0.011471213772892952,0.00044845533557236195,-0.03572803735733032,0.008408864960074425,0.005150725599378347,0.0027136774733662605,-0.015526692382991314,0.019093889743089676,0.023867793381214142,0.002074668649584055,0.015213425271213055,-0.03485359251499176,-0.01568167470395565,0.0019035213626921177,-0.01312799472361803,-0.0005730802076868713,0.00797811709344387,-0.004188996739685535,-0.0004184977733530104,-0.017942678183317184,0.034504521638154984,0.02762117236852646,-0.01790756545960903,0.0009181434870697558,-0.03011685609817505,0.014275695197284222,0.015353744849562645,-0.031707435846328735,-0.014739478938281536,0.016829000785946846,0.017712701112031937,0.007461728528141975,0.01610303856432438,0.06119834631681442,0.026246709749102592,-0.018305666744709015,0.01999809592962265,0.020704712718725204,0.0006690274458378553,0.006749357562512159,-0.02893499657511711,0.030817503109574318,0.012018632143735886,-0.00804202351719141,-0.018375836312770844,0.010928541421890259,0.010707253590226173,0.013002028688788414,-0.002033184515312314,-0.005340408068150282,0.03387705981731415,-0.022892260923981667,0.014678869396448135,0.013537639752030373,-0.008713518269360065,-0.02593693695962429,0.07454174757003784,-0.024845819920301437,0.02834264747798443,-0.02971474640071392,-0.011557130143046379,-0.008002926595509052,-0.019368749111890793,0.028180744498968124,0.020025016739964485,0.00347648817114532,-0.024566639214754105,0.010174321942031384,-0.0041660708375275135,-0.032732993364334106,0.020300105214118958,-0.015345271676778793,-0.014129326678812504,-0.018271800130605698,-0.051289476454257965,0.010386044159531593,-0.009736659936606884,0.058951228857040405,-0.019437629729509354,0.010607936419546604,-0.035290662199258804,-0.018360501155257225,0.00655018538236618,0.006995379459112883,-0.0038300538435578346,0.025012381374835968,-0.016321701928973198,0.030617311596870422,0.009497919119894505,0.03148818388581276,-0.0013535958714783192,0.04819456487894058,-0.029439838603138924,-0.01625029742717743,0.02551189251244068,-0.012317327782511711,0.00624461006373167,0.0006637125625275075,0.02077995426952839,-0.020640326663851738,0.019393613561987877,0.010709955357015133,0.0266432985663414,0.007108711171895266,-0.004445891827344894,-0.00882738083600998,0.00688306475058198,0.041268378496170044,-0.007304352708160877,-0.03441083803772926,0.008169202134013176,0.011502728797495365,-0.007084041368216276,-0.020305443555116653,-0.025182053446769714,0.017318835482001305,0.033297374844551086,0.01863124966621399,0.021403657272458076,0.00891103409230709,0.006325514055788517,-0.018800152465701103,-0.010625758208334446,-0.026598088443279266,0.0007023508660495281,-0.01763393171131611,0.016984479501843452,-0.004765206947922707,-0.02932245470583439,-0.029315751045942307,-0.028604162856936455,0.001683940296061337,-0.002616151934489608,-0.0017021234380081296,-0.0058658914640545845,-0.01610795594751835,-0.013643622398376465,-0.015004831366240978,-0.02692246250808239,0.036434344947338104,0.050759829580783844,-0.02177542634308338,-0.02895553968846798,0.004390752874314785,-0.0071328007616102695,-0.0051391092129051685,-0.021695662289857864,-0.02772023156285286,0.0008286261581815779,0.005659732036292553,-0.006769032217562199,-0.013968240469694138,-0.026159070432186127,0.025013821199536324,-0.011028788983821869,0.015586946159601212,0.006575345527380705,-0.007776331622153521,-0.019144093617796898,0.00411997688934207,0.007583860773593187,-0.025160323828458786,0.0077173043973743916,-0.007576176896691322,0.003944530617445707,-0.029183954000473022,-0.01147263590246439,-0.016668498516082764,-0.011516440659761429,0.003953504376113415,0.00978126097470522,0.029793299734592438,-0.015394643880426884,0.010064088739454746,-0.021008552983403206,-0.01573927327990532,-0.030099501833319664,0.0003759782121051103,0.0006849328055977821,0.0389038547873497,-0.02146504633128643,-0.018647141754627228,-0.01572437584400177,-0.007592550944536924,-0.0015588185051456094,0.002662237733602524,0.007236766628921032,0.01272666733711958,0.026706160977482796,0.05235762149095535,-0.012421092949807644,0.00013509750715456903,0.04476403817534447,-0.009771216660737991,-0.02118249610066414,0.04454591125249863,-0.017028428614139557,-0.01195906475186348,0.005678544286638498,0.00836592260748148,0.01189937349408865,0.012730962596833706,-0.005630319472402334,0.01452713180333376,-0.013275792822241783,-0.018467117100954056,0.014105362817645073,-0.027940034866333008,0.016660083085298538,0.004231704398989677,-0.01585736684501171,0.02311568148434162,-0.012323492206633091,0.030989425256848335,-0.0017990367487072945,0.005157525651156902,0.011860912665724754,-0.003062732517719269,-0.0021780328825116158,0.027445411309599876,0.0037140955682843924,-0.027354771271348,0.01905607432126999,-0.021779637783765793,0.004786330740898848,-0.014433914795517921,-0.000988819869235158,-0.025424044579267502,-0.02375912293791771,0.0002571911318227649,-0.01142153050750494,0.00633115042001009,-0.03022611141204834,-0.006386420223861933,0.015073674730956554,0.021177120506763458,0.0010647213784977794,0.02805747091770172,-0.012359117157757282,-0.00851079635322094,0.008847501128911972,0.03417434170842171,-0.0038769945967942476,0.028228238224983215,-0.03615333139896393,-0.010117767378687859,0.03377214074134827,-0.037624865770339966,0.00220870622433722,0.00726681062951684,0.0031843953765928745,-0.00023749811225570738,-0.014478919096291065,-0.019637541845440865,-0.017522139474749565,-0.02418166771531105,-0.02212562784552574,0.05112030357122421,-0.025338690727949142,0.034731388092041016,0.03619275614619255,-0.018059806898236275,-0.012308582663536072,0.02949472703039646,-0.005662223789840937,-0.034676920622587204,-0.019521132111549377,-0.0032274748664349318,0.010125264525413513,0.016992516815662384,0.02023046649992466,0.025654837489128113,-0.0035119461826980114,-0.01830037496984005,0.0412350594997406,-0.025699708610773087,0.3945489227771759,0.0022388233337551355,0.006623091641813517,0.02957732230424881,-0.02066163346171379,0.00981396995484829,-0.008598527871072292,0.013277895748615265,0.020120883360505104,-0.006477731745690107,0.030418142676353455,-0.007340218871831894,-0.012548127211630344,0.023639768362045288,0.010163206607103348,-0.034796204417943954,0.0006370857008732855,-0.01274147629737854,-0.02447603829205036,-0.03882390633225441,-0.031244494020938873,-0.026461591944098473,-0.0022103397641330957,-0.019572583958506584,-0.016439631581306458,0.015622923150658607,-0.015521136112511158,0.03595787286758423,-0.0048918514512479305,-0.002181324642151594,0.016870345920324326,0.019937478005886078,0.041617486625909805,-0.005293984431773424,0.007304266560822725,-0.023329349234700203,-0.0038777447771281004,0.0014489589957520366,-0.009314816445112228,-0.022992243990302086,0.003761240281164646,0.0337349958717823,0.0223101619631052,0.013082288205623627,-0.026393432170152664,0.006819479167461395,-0.021051079034805298,0.010396712459623814,-0.034099213778972626,0.022815337404608727,-0.0009314282215200365,0.02074781432747841,0.0007524539832957089,0.02071470208466053,0.009940960444509983,-0.012605628930032253,-0.007611552253365517,0.034557487815618515,-0.015403050929307938,0.0068310028873384,-0.026989562436938286,-0.004270288627594709,-0.020345613360404968,0.01306311134248972,-0.013799809850752354,-0.029911305755376816,-0.009493411518633366,-0.011653181165456772,-0.009218117222189903,-0.0007806898793205619,0.01938917674124241,0.001906267600134015,-0.019101956859230995,-0.007550311740487814,-0.012143472209572792,-0.04367460682988167,0.025638077408075333,-0.014629567973315716,0.028194574639201164,-0.01728118024766445,0.0221963319927454,0.019204938784241676,0.03432143107056618,-0.02657555416226387,0.002004743553698063,-0.020791711285710335,-0.010341069661080837,-0.04619452357292175,-0.005083365831524134,-0.00990467518568039,0.002460786607116461,0.010507918894290924,0.016275444999337196,-0.06338823586702347,-0.007412796374410391,0.06120302900671959,0.010436675511300564,0.0034267818555235863,0.02876623161137104,-0.0007090673316270113,-0.008585995063185692,-0.011970008723437786,0.005763562396168709,-0.024609150364995003,0.02850452810525894,-0.022673225030303,0.01191666629165411,-0.04924258962273598,-0.04102671891450882,0.02912851609289646,-0.016620686277747154,-0.03335510939359665,-0.01148713193833828,-0.010274672880768776,0.009239780716598034,-0.004761083517223597,0.02867140993475914,0.014553618617355824,-0.021684858947992325,0.018985014408826828,0.014055677689611912,0.026981523260474205,-0.004754500463604927,-0.00038642145227640867,-0.013451570644974709,-0.016589971259236336,0.022141065448522568,-0.004539037588983774,0.03152543306350708,0.01208160538226366,-0.024780472740530968,0.02048008143901825,0.006478175055235624,0.03647821769118309,-0.02186744473874569,0.033406343311071396,0.021772053092718124,-0.0025682891719043255,0.04414818063378334,0.00921628624200821,0.008988365530967712,0.007342945784330368,-0.007830486632883549,0.03257960081100464,-0.012104877270758152,0.009669927880167961,-0.0064807008020579815,-0.009303362108767033,0.007799571845680475,0.01879187859594822,0.013339410535991192,0.01986137218773365,0.0008062629494816065,-0.2452234923839569,0.0016491387505084276,-0.04547182098031044,-0.0028053668793290854,0.009465817362070084,-0.021054264158010483,0.008472413755953312,0.0708577111363411,-0.009206854738295078,0.0014748502289876342,0.007380124647170305,0.02171417511999607,-0.003809248097240925,0.013634135015308857,-0.0012557351728901267,0.01094928290694952,0.009313702583312988,-0.02872377634048462,0.0031375717371702194,-0.01976471208035946,0.0020988467149436474,0.003526208456605673,0.01742757111787796,-0.009931430220603943,-0.011483557522296906,-0.016786184161901474,0.003717949381098151,-0.007670466788113117,-0.0036177970468997955,-0.0075657786801457405,-0.014233733527362347,-0.016591597348451614,0.019823554903268814,-0.012458041310310364,-0.012151968665421009,0.012052053585648537,0.023169804364442825,0.007043288554996252,-0.06283272057771683,0.009973570704460144,0.04159274697303772,-0.0027213972061872482,-0.026247259229421616,-0.0005786009132862091,0.005472692660987377,0.010092354379594326,-0.0004666309105232358,0.0027546195778995752,-0.021309800446033478,0.03282300382852554,0.009572925977408886,-0.020475301891565323,-0.02145480178296566,-0.027329636737704277,-0.027623644098639488,0.00212560105137527,0.021240131929516792,0.004549054894596338,-0.009809929877519608,-0.01735849492251873,-0.00037379004061222076,0.02335747703909874,-0.007545780390501022,0.0291381124407053,0.026592694222927094,-0.004482008051127195,0.030557220801711082,-0.020553363487124443,-0.01994723454117775,-0.0064659821800887585,0.02962430939078331,-0.0008922646520659328,-0.053251899778842926,0.009864742867648602,0.020337197929620743,-0.003998814150691032,-0.004185308702290058,6.866786134196445e-05,0.014037802815437317,0.027247929945588112,0.048447780311107635,-0.01015166100114584,0.02130924165248871,-0.0233079232275486,0.02607346512377262,0.040281493216753006,-0.007543081417679787,-0.035924218595027924,0.00569900730624795,-0.008039957843720913,-0.002823712769895792,-0.002803286537528038,-0.006118007004261017,0.022434748709201813,0.013933056965470314,0.002936281030997634,-0.013914977200329304,0.023689141497015953,0.007962143048644066,-0.011495125479996204,-0.02081713080406189,0.01413489505648613,0.012353007681667805,-0.015297238714993,-0.02939029410481453,-0.015370906330645084,0.015056128613650799,0.014590793289244175,0.026387853547930717,-0.00512867234647274,-0.008124465122818947,-0.015975723043084145,0.02681264467537403,-0.010784736834466457,0.01441181544214487,0.03277438506484032,0.010350184515118599,0.009400716051459312,-0.02116876281797886,0.0018132136901840568,-0.03765971586108208,0.0052503058686852455,0.03420824930071831,0.014777399599552155,-0.016579443588852882,-0.03330729156732559,-0.013327029533684254,0.0005119363777339458,-0.020840656012296677,-0.004507529083639383,-0.011724368669092655,-0.03594064712524414,0.0015368282329291105,0.009197228588163853,6.866830517537892e-05,0.0282279159873724,-0.052910640835762024,-0.040126197040081024,-0.029475118964910507,0.005734991282224655,-0.004573932383209467,-0.014002541080117226,0.002804175252094865,0.015738246962428093,0.026456352323293686,-0.011875574477016926,-0.030365511775016785,-0.008460002951323986,0.0012545310892164707,0.010170886293053627,0.037671059370040894,-0.014676935039460659,0.004320601467043161,-0.001990457298234105,-0.008405500091612339,0.009865849278867245,-0.00018689411808736622,0.00172912678681314,0.024112839251756668,-0.018725017085671425,0.02372124046087265,-0.029248137027025223,0.01865905337035656,0.022101448848843575,-0.026333123445510864,-0.03643358498811722,-0.009696442633867264,0.0064179981127381325,-0.02068602852523327,-0.0005584901664406061,-0.009454110637307167,0.03605068847537041,0.0053380923345685005,0.03280332684516907,-0.03048955276608467,0.0333951935172081,-0.013537755236029625,0.011699914000928402,0.00550722936168313,-0.003158178646117449,-0.0023985630832612514,0.0008453089976683259,-0.015513830818235874,-0.01306804921478033,0.014389708638191223,0.00964529998600483,-0.025586921721696854,0.04138404130935669,0.01857851818203926,0.003144270507618785,-0.014571740292012691,0.0008016453357413411,-0.02522323466837406,-0.021038807928562164,0.0260340366512537,-0.014219886623322964,0.02239949256181717,-0.03005603700876236,-0.0002968251646962017,0.04302574694156647,-0.011221950873732567,0.013686021789908409,0.011945300735533237,-0.021525831893086433,0.016680726781487465,-0.009338047355413437,-0.0013850744580850005,-0.004670086316764355,-0.01883036084473133,-0.021457115188241005,0.016184847801923752,-0.02029011771082878,0.012913129292428493,0.03236585855484009,-0.025957178324460983,-0.019733259454369545,0.0014983630971983075,-0.0009224032983183861,0.018271666020154953,-0.016837170347571373,-0.03021763078868389,0.024666650220751762,-0.005840935278683901,-0.0002922657295130193,0.01992247812449932,0.0025119774509221315,0.00942355114966631,-0.024607347324490547,0.0175350159406662,0.008702065795660019,-0.0030363621190190315,0.029419153928756714,0.02159796841442585,-0.02282441034913063,-0.007350334897637367,-0.04385586455464363,0.025481482967734337,-0.0009410607162863016,-0.02334352768957615,0.03386714309453964,-0.00873310212045908,-0.026216188445687294,-0.006248565390706062,-0.0017794203013181686,-0.007381954696029425,0.012409504503011703,0.009813938289880753,0.02856520004570484,-0.016347434371709824,0.009081486612558365,-0.029968498274683952,-0.03468979150056839,0.021744443103671074,-0.017760567367076874,-0.023141345009207726,0.01621149852871895,-0.022202035412192345,0.013410212472081184,-0.04080522432923317,-0.02034514583647251,0.01637434773147106,-0.01152025070041418,0.046551160514354706,-0.028310056775808334,-0.01566505618393421,0.008510574698448181,-0.004622760694473982,-0.04889841377735138,-0.007965917699038982,0.05206499621272087,0.01185608096420765,0.009846740402281284,0.014891531318426132,-0.058800600469112396,0.008623807691037655,-0.02387350983917713,0.028119025751948357,-0.030440298840403557,-0.0022277296520769596,0.008353536017239094,0.022365428507328033,0.04501890018582344,0.006614751648157835,0.009785963222384453,-0.0009170128032565117,-0.2600760757923126,-0.016080114990472794,-0.045600809156894684,0.37184008955955505,0.00042362671229057014,-0.04083505645394325,-0.0060874950140714645,0.004676590207964182,-0.0015948755899444222,0.006051524542272091,0.009526277892291546,-0.024084242060780525,0.0027569576632231474,0.0031010331586003304,-0.01725873164832592,0.012725659646093845,0.020268095657229424,-0.018510079011321068,-0.02837648242712021,0.03285406157374382,-0.04530850052833557,0.013998305425047874,0.006253780797123909,-0.008995305746793747,-0.03807469829916954,-0.005177452228963375,0.0016503955703228712,0.002722939010709524,0.01649770513176918,-0.012529625557363033,-0.011780786328017712,-0.04194258153438568,0.2932291626930237,-0.005160861648619175,-0.025871288031339645,-0.015019666403532028,0.010148548521101475,-0.031369827687740326,0.010749276727437973,0.012993646785616875,-0.032632481306791306,0.008153950795531273,-0.028490586206316948,0.013572984375059605,0.014727172441780567,0.02997964806854725,0.03599761798977852,-0.011955379508435726,-0.0023449331056326628,0.007874103263020515,0.0031892163679003716,-0.0630422905087471,-0.045875441282987595,0.0045535508543252945,-0.006551229860633612,0.014457279816269875,0.02571825310587883,0.0025780280120670795,0.004330016206949949,-0.034172870218753815,-0.0011205438058823347,-0.039298295974731445,0.013649556785821915,-0.019661420956254005,0.007333013229072094,0.0033112831879407167,-0.00907302089035511,0.05523563548922539,0.06825254112482071,-0.029133841395378113,0.007138827815651894,0.0018581028562039137,-0.0027824908029288054,-0.018279701471328735,0.013761738315224648,0.007517494726926088,-0.05192633718252182,0.019648293033242226,-0.0065101259388029575,0.0036486471071839333,-0.011046899482607841,-0.03020889312028885,-0.008332557044923306,-0.023126818239688873,-0.02244790829718113,-0.01924694888293743,0.03122282400727272,0.005073893815279007,-0.01787622645497322,0.021307969465851784,0.014510958455502987,0.010585499927401543,0.014507881365716457,-0.002168098231777549,0.014450464397668839,0.013872889801859856,-0.005848449654877186,0.006843095179647207,-0.01401022169739008,0.007842911407351494,-0.0068886177614331245,-0.015383785590529442,-0.00011942457786062732,-0.0436474084854126,-0.01509405579417944,-0.02598675526678562,0.00272564310580492,0.0038308147341012955,0.004427497275173664,-0.00624738959595561,0.009107514284551144,-0.020619701594114304,0.010628441348671913,0.005285948980599642,0.014739288948476315,0.005811752285808325,0.0026360531337559223,-0.018963048234581947,-0.009475690312683582,0.05785176157951355,-0.01737930439412594,-0.02488931454718113,-0.0399969145655632,0.003923316486179829,0.018002383410930634,-0.02504521608352661,-0.023813752457499504,-0.017232278361916542,0.00795676652342081,-0.001794273965060711,-0.020255418494343758,-0.000540823966730386,-0.001777304452843964,0.006025017239153385,-0.0427837073802948,0.019155099987983704,-0.023230666294693947,0.02336927130818367,-0.007677625399082899,-0.01661953143775463,0.03562818095088005,0.019376536831259727,-0.03837037459015846,-0.04313799738883972,0.015176735818386078,0.008284813724458218,0.009473254904150963,-0.03282340243458748,-0.00046932106488384306,0.018362799659371376,0.01757805608212948,0.015880797058343887,-0.0002934168151114136,0.017278384417295456,0.01328978594392538,-0.02181888371706009,-0.004277618136256933,0.0011071941116824746,-0.005517182871699333,0.026351038366556168,0.028246188536286354,0.0002695833973120898,0.045881159603595734,0.026812918484210968,-0.04657647758722305,0.02684716135263443,-0.010451615788042545,-0.0068083698861300945,0.019908539950847626,-0.022675009444355965,0.01397355180233717,-0.006318001076579094,-0.03992696851491928,0.013012533076107502,0.029743270948529243,-0.004279462620615959,-0.020512741059064865,-0.006946631707251072,-0.0026941944379359484,0.010775614529848099,0.018565719947218895,-0.003007622202858329,-0.013308489695191383,0.021564042195677757,-0.002796269953250885,-0.010052744299173355,-0.02053438313305378,0.005811529234051704,0.00324440049007535,0.05007342994213104,0.0023459268268197775,0.026071349158883095,0.0456637442111969,-0.02146206982433796,0.02050071582198143,-0.01621152274310589,-0.0037330505438148975,0.0015077710850164294,0.004053729120641947,-0.012518640607595444,0.005028035957366228,0.019160188734531403,0.027091413736343384,-0.007997505366802216,-0.06230811029672623,-0.03779023885726929,-0.0016847921069711447,0.004708516411483288,-0.016163809224963188,-0.006259354762732983,0.03144799917936325],\"object\":\"embedding\"},{\"index\":1,\"embedding\":[-0.025758948177099228,0.012529470957815647,-0.0036885219160467386,0.013843787834048271,0.037662822753190994,-0.002338479971513152,0.014036386273801327,-0.02571997605264187,-0.02729753963649273,0.023403557017445564,-0.025056499987840652,0.006940447259694338,0.005110990721732378,0.002051352057605982,0.009365430101752281,-0.004972649272531271,-0.008563087321817875,-0.04744229465723038,-0.007888401858508587,-0.0362674854695797,-0.04641663655638695,-0.006453453563153744,-0.0054247332736849785,-0.007706332020461559,-0.0316537506878376,0.030296489596366882,-0.03128792345523834,-0.052762240171432495,0.02158757671713829,-0.021084534004330635,0.04210008308291435,-0.009039834141731262,0.020920418202877045,0.011608952656388283,-0.027386372908949852,0.0006602472276426852,0.020277611911296844,-0.006371058989316225,-0.00742132356390357,-4.1149793105432764e-05,-0.026434097439050674,0.02487286366522312,0.005916973575949669,0.004006581846624613,0.013903914019465446,0.03047511726617813,0.010963551700115204,-0.0004852090496569872,-0.008918477222323418,0.0024200777988880873,-0.005843935534358025,-0.009319420903921127,-0.03840300813317299,-0.00996500626206398,0.0029545770958065987,-0.011163728311657906,0.01174638420343399,-0.01566792093217373,-0.02538173645734787,-0.017242424190044403,-0.02570188045501709,0.008348853327333927,-0.024366218596696854,0.012855932116508484,0.027172120288014412,-0.01550713274627924,-0.008245732635259628,-0.004089049994945526,0.011690033599734306,0.011521993204951286,-0.03591177612543106,-0.012615636922419071,-0.02868751250207424,-0.015085900202393532,-0.038233257830142975,-0.016448313370347023,0.037549398839473724,-1.5810148397576995e-05,-0.020648358389735222,-0.016178883612155914,0.02934243343770504,0.015080793760716915,0.034065019339323044,0.007480921223759651,0.00841328501701355,-0.04200136661529541,-0.024020958691835403,-0.0023532297927886248,-0.02087455242872238,-0.05933527648448944,-0.029153307899832726,-0.020886346697807312,0.0040543656796216965,0.03735136613249779,0.018253160640597343,-0.010190599597990513,-0.006286981515586376,0.01285966020077467,0.018434688448905945,-0.01497467141598463,0.032310742884874344,-0.07860245555639267,0.0018877912079915404,0.028160221874713898,-0.012849449180066586,-0.0030984075274318457,0.007885973900556564,0.013520868495106697,-0.023572495207190514,0.002127881860360503,0.011660636402666569,-0.006299937609583139,0.007196198217570782,-0.03473691642284393,-0.0029503554105758667,-0.027778655290603638,-0.03009990230202675,0.014822233468294144,0.004842456895858049,-0.019102793186903,0.020562220364809036,0.01187469158321619,-0.008432718925178051,0.041573893278837204,0.037788111716508865,0.008579518646001816,0.025191793218255043,-0.05494324117898941,0.02662293054163456,0.008876971900463104,0.004079464357346296,-0.03443434461951256,0.016408467665314674,0.01451954711228609,0.03553950786590576,0.0038041200023144484,0.0015072039095684886,0.04059716686606407,0.0009355077636428177,-0.017191415652632713,-0.004116380121558905,-0.03172164037823677,0.06326121836900711,-0.042410608381032944,-0.01892443373799324,0.000301255815429613,0.014980636537075043,0.008668444119393826,-0.006467134226113558,0.01782114803791046,-0.00831829197704792,0.044133491814136505,-0.0182215366512537,-0.04831685125827789,0.03808142989873886,0.040061261504888535,-0.038982439786195755,0.025533616542816162,-0.01174217090010643,0.019772527739405632,0.015522862784564495,0.0030584626365453005,-0.004236175678670406,0.021114911884069443,0.003085960168391466,0.0017570361960679293,0.001260936027392745,0.01698777638375759,-0.037901267409324646,0.04388272017240524,-0.005779674742370844,-0.006999421864748001,-0.0004813986597582698,-0.007257245946675539,-0.009489151649177074,-0.015425185672938824,0.03870683163404465,-0.001314341789111495,-0.01330617442727089,-0.06468790024518967,-0.03725465387105942,-0.03481857851147652,-0.029993493109941483,0.001782639417797327,0.014728524722158909,-0.023663615807890892,0.003847962012514472,0.016591563820838928,0.009149174205958843,0.013524849899113178,0.014362265355885029,-0.05420935899019241,0.015831908211112022,0.02472783997654915,0.011116784065961838,-0.02629142627120018,-0.006261439062654972,0.007880052551627159,-0.008645856752991676,0.0014268093509599566,0.014113090001046658,0.004895458929240704,0.010227800346910954,-0.03291991353034973,0.014134642668068409,-0.00833825021982193,0.021962054073810577,-0.027159973978996277,0.0009441711590625346,-0.023185022175312042,-0.0014423648826777935,-0.02066347748041153,-0.05037784203886986,0.011989530175924301,0.019780971109867096,0.02544422633945942,0.011617295444011688,0.05003154277801514,-0.06384427100419998,0.03010302409529686,0.010974934324622154,-0.02524236962199211,-0.04694395512342453,0.04317052662372589,-0.014662305824458599,-0.011110682971775532,-0.013760543428361416,0.032241374254226685,-0.021328603848814964,-0.005873484071344137,0.020389661192893982,-0.0020480805542320013,-0.036380380392074585,-0.003262253012508154,0.022140566259622574,0.014284894801676273,0.02966931276023388,-0.014425903558731079,-0.007815143093466759,-0.03838087618350983,-0.01601572334766388,-0.008720634505152702,0.00740523636341095,-0.010962577536702156,-0.01864113286137581,-0.03902296721935272,0.0023240335285663605,-0.03017006255686283,-0.010889565572142601,-0.00671680923551321,0.0002762922376859933,-0.0018828880274668336,-0.0048309871926903725,0.0182066410779953,0.034271206706762314,0.051186565309762955,-0.00884862057864666,-0.01352836936712265,0.0046797567047178745,0.01883058063685894,-0.013339086435735226,-0.0057356832548975945,0.031059427186846733,0.013298808597028255,0.01392305176705122,-0.01582452654838562,-0.007665578275918961,-0.04605823755264282,0.01951473578810692,-0.0003280778764747083,-0.017294079065322876,-0.007146188989281654,0.011400898918509483,0.02734067291021347,0.006338881328701973,-0.006668481510132551,0.05188588425517082,0.002703281817957759,-0.02523582987487316,-0.010897639207541943,0.01924540475010872,-0.0009212141158059239,-0.012981880456209183,0.028238492086529732,0.03482842445373535,0.027267372235655785,0.03475086763501167,-0.021356694400310516,-0.0003882267337758094,-0.02259952947497368,0.031467314809560776,-0.011028904467821121,-0.005698627792298794,0.010554694570600986,0.04053696244955063,-0.001511637819930911,-0.061206888407468796,-0.0029146643355488777,0.01215550396591425,0.006889894139021635,0.044212453067302704,-0.030752873048186302,0.004620553459972143,-0.00023938347294460982,-0.024459535256028175,-0.009339572861790657,-0.04982149228453636,0.033970270305871964,0.021464470773935318,-0.006805697921663523,0.026049407199025154,-0.01640831120312214,0.0077367620542645454,-0.04079772159457207,-0.016999192535877228,0.030637428164482117,0.00034822430461645126,0.020782500505447388,0.025917865335941315,0.045175231993198395,0.004338984377682209,0.004516663961112499,-0.03431222215294838,-0.0018257899209856987,-0.010955095291137695,0.004570914898067713,0.3870064914226532,0.026963500306010246,-0.035104792565107346,0.011724506504833698,-0.013705889694392681,-0.0509878471493721,-0.02112174779176712,0.008177575655281544,0.01993235945701599,-0.014281210489571095,0.017260506749153137,-0.045702073723077774,0.006509148050099611,0.0032414980232715607,0.005621928256005049,0.034762829542160034,-0.0007084138342179358,-0.020915701985359192,0.005990590900182724,0.021028028801083565,-0.045931894332170486,0.010121088474988937,0.0044517978094518185,-0.03989334776997566,-0.008309227414429188,-0.02713819220662117,-0.00228235125541687,-0.01568441092967987,-0.02216830477118492,-0.004598020110279322,-0.010677327401936054,-0.009580195881426334,-0.003543584141880274,0.03834174945950508,0.01644299365580082,-0.03651408851146698,0.0067563727498054504,-0.015657847747206688,0.0012091726530343294,-0.033857397735118866,-0.012780657038092613,0.005480979103595018,-0.03239237889647484,0.013967982493340969,0.03662465885281563,-0.03532421588897705,0.01288940105587244,-0.02662917785346508,-0.04374580830335617,-0.015550529584288597,-0.006989222951233387,0.011201686225831509,0.06042121350765228,0.007955683395266533,-0.003084216732531786,-0.011056994087994099,0.0037636784836649895,0.007206861395388842,0.027355395257472992,-0.00904571358114481,-0.04939938336610794,0.019223814830183983,0.04885827749967575,0.004720363765954971,-0.03998367488384247,-0.04003869742155075,-0.010004842653870583,-0.009810425341129303,-0.011637724004685879,0.01382185984402895,0.0011330196866765618,0.02921152301132679,0.007083197124302387,0.015879228711128235,-0.011928517371416092,-0.011234760284423828,0.0018301004311069846,-0.024675864726305008,-0.008007164113223553,0.05101221799850464,-0.01141275092959404,0.018010249361395836,-0.02040042355656624,-0.017251471057534218,-0.015094183385372162,0.024670656770467758,-0.007679453119635582,-0.0023301681503653526,0.01593824103474617,-0.017610495910048485,0.0392007976770401,0.039871010929346085,0.038008034229278564,-0.034386586397886276,0.03700798749923706,-0.02587129920721054,-0.0024405464064329863,-0.024259598925709724,0.007090513128787279,0.0037076836451888084,0.038092534989118576,0.002115082461386919,-0.0011044091079384089,-0.015718095004558563,0.0022297045215964317,0.025628330186009407,0.03330320864915848,0.011225000023841858,-0.01548957172781229,0.01112839300185442,-0.0005184286274015903,0.012068329378962517,-0.0018296351190656424,-0.0198866818100214,0.021056117489933968,0.042734939604997635,0.011327057145535946,-0.010904025286436081,-0.004531412851065397,0.0400407649576664,0.005563332233577967,-0.012673179619014263,0.0009829836199060082,-0.008927428163588047,-0.047734569758176804,-0.0014640727313235402,0.014513886533677578,0.014106206595897675,-0.02875625342130661,0.027171747758984566,0.05241311714053154,0.02950497902929783,-0.030510742217302322,-0.006987013854086399,-0.006407857406884432,0.03180888295173645,0.004917642567306757,0.010594794526696205,-0.025200797244906425,-0.018448997288942337,-0.0018395927036181092,-0.010068191215395927,-0.06761447340250015,0.03835010156035423,0.02366061508655548,-0.014682404696941376,0.022400829941034317,-0.05012622848153114,-0.02561725489795208,0.027008507400751114,0.0015325259882956743,0.05601796507835388,-0.05630066990852356,0.02009653113782406,-0.02773994207382202,-0.04563506320118904,-0.01889072358608246,-0.05071859434247017,-0.008834474720060825,0.0068897102028131485,-0.015094980597496033,-0.0444563664495945,-0.03613041713833809,-0.02099127508699894,-0.03687067702412605,-0.044355358928442,-0.010367853567004204,0.028472913429141045,0.01250379253178835,0.016446998342871666,-0.031522832810878754,0.0023249490186572075,0.018918611109256744,0.0017459378577768803,-0.019281113520264626,0.019197694957256317,-0.007465716917067766,0.04378819465637207,-0.012430009432137012,-0.007315130904316902,0.005798067897558212,0.026136348024010658,-0.05674370378255844,-0.02571779303252697,-0.011887992732226849,-0.01981491968035698,-0.010861019603908062,0.03535069525241852,-0.028128519654273987,0.004035569727420807,0.05296939238905907,-0.0008354758028872311,0.010797429829835892,0.027159707620739937,0.024418288841843605,0.01101562287658453,0.020692870020866394,0.006610993295907974,0.026814602315425873,0.04471895843744278,0.01140670944005251,0.0178330447524786,0.020213549956679344,-0.0070716277696192265,-0.04305743798613548,0.011643058620393276,-0.005915377289056778,0.03349792957305908,-0.026610828936100006,-0.004258973523974419,-0.02179308421909809,0.042003829032182693,-0.025853054597973824,-0.012385998852550983,0.002028633374720812,-0.022666480392217636,0.01318320445716381,0.01882292702794075,0.031063402071595192,-0.03805585950613022,0.035645823925733566,0.002151072956621647,0.007318989373743534,0.016153370961546898,0.04089585691690445,-0.023799102753400803,0.013497691601514816,-0.026399897411465645,-0.049984849989414215,-0.017662756145000458,0.03448191657662392,-0.008758445270359516,0.008639284409582615,-0.013155733235180378,-0.018071550875902176,0.019655385985970497,-0.013738088309764862,0.04678110033273697,0.027242839336395264,0.0018837314564734697,-0.030672207474708557,0.011950843036174774,0.029605276882648468,0.017911622300744057,-0.015331711620092392,-0.014341580681502819,0.024940475821495056,0.009718193672597408,0.00183775182813406,0.031651679426431656,0.027826672419905663,0.007948522455990314,-0.005205032881349325,0.02395801618695259,-0.03033900447189808,-0.026571951806545258,-0.03070627897977829,-0.011602534912526608,0.019729234278202057,0.018046380952000618,-0.019431443884968758,-0.00806436873972416,-0.007017528638243675,-0.0008441002573817968,0.008560731075704098,-0.01921754702925682,-0.005517745856195688,-0.007153145503252745,0.04134321212768555,0.0026307522784918547,-0.02331283688545227,-0.0027744488324970007,0.009677980095148087,0.021427741274237633,-0.004601431544870138,-0.00012251334555912763,0.028399625793099403,-0.006959706079214811,-0.01915843039751053,0.03018774650990963,-0.03351955860853195,0.06670859456062317,-0.03129318356513977,-0.009517683647572994,0.049102671444416046,-0.010580141097307205,0.01756119914352894,0.02052275277674198,-0.00034793917438946664,-0.05726324021816254,-0.021166298538446426,-0.005764365196228027,0.028409039601683617,-0.004559235647320747,-0.036898449063301086,0.016247611492872238,0.04555016756057739,-0.025495193898677826,0.016971617937088013,-0.0012656591134145856,-0.028013968840241432,0.02536524273455143,-0.016685131937265396,0.003650881117209792,-0.001630979822948575,-0.01653134636580944,-0.006658824160695076,0.03132772073149681,0.011927434243261814,-0.026557045057415962,-0.19315965473651886,0.006070801988244057,-0.018877672031521797,0.015526301227509975,0.04478176310658455,-0.0008945377776399255,0.013174171559512615,0.023236852139234543,0.001876345369964838,-0.009473146870732307,-0.0458466038107872,0.028369978070259094,-0.03175703436136246,-0.030467139557003975,0.05286461114883423,0.006947376299649477,-0.018354244530200958,0.013310008682310581,-0.010109184309840202,-0.014129184186458588,-0.0019312937511131167,-0.014013266190886497,-0.010930518619716167,-0.039712607860565186,-0.0027366008143872023,0.022063201293349266,0.041075583547353745,0.009186118841171265,0.018548376858234406,-0.021595891565084457,-0.00893919263035059,-0.007997609674930573,-0.011303680017590523,-0.00778892682865262,0.003854784881696105,0.0269649475812912,0.003425600938498974,0.0011551182251423597,0.014906526543200016,-0.0268486887216568,0.015963394194841385,-0.037813760340213776,0.059832315891981125,-0.012539348565042019,-0.024167275056242943,0.010521982796490192,-0.01648557558655739,0.004499785602092743,-0.0023810069542378187,0.006643267348408699,-0.001206551562063396,-0.009098317474126816,0.004006785340607166,0.016244588419795036,0.022345518693327904,-0.003811529139056802,0.00012655694445129484,-0.007302777376025915,-0.03910021483898163,-0.011724776588380337,-0.022477485239505768,0.012693922966718674,-0.014236903749406338,0.012715728022158146,0.013103259727358818,0.0066465092822909355,0.020985668525099754,0.013942337594926357,0.0061926888301968575,-0.03711562603712082,-0.04018467292189598,0.004651527386158705,0.028886903077363968,-0.003899460658431053,0.014873364940285683,-0.051289740949869156,0.023774413391947746,0.025767862796783447,-0.008617984130978584,-0.01755518466234207,0.04933374375104904,0.023852013051509857,-0.02684945799410343,0.0013523841043934226,-0.015365022234618664,-0.059762801975011826,0.02864656411111355,0.026391534134745598,0.03813229128718376,-0.012687886133790016,0.028539935126900673,0.08154822140932083,0.010055551305413246,0.011949121952056885,-0.013092302717268467,-0.026565100997686386,-0.012263688258826733,-0.03514808788895607,-0.008544047363102436,0.04167114570736885,0.0164664164185524,-0.021921271458268166,0.030826108530163765,-0.0005666292272508144,0.030772607773542404,0.028927890583872795,-0.026523489505052567,-0.040638573467731476,0.010861112736165524,0.01859780214726925,-0.02758798561990261,0.00662233354523778,-6.221868534339592e-05,-0.2581826150417328,-0.007566799875348806,0.023047540336847305,0.05015581473708153,0.022923756390810013,0.011462699621915817,-0.006452121306210756,0.01042680349200964,-0.045284438878297806,-0.03936319053173065,0.025846514850854874,0.01967019960284233,0.01158192753791809,-0.002621933352202177,-0.004696441814303398,-0.007103266194462776,-0.027700381353497505,-0.01727375201880932,-0.054769087582826614,-0.0009374514338560402,-0.0020084483548998833,-0.029182057827711105,-0.02928612194955349,-0.013479417189955711,0.014092243276536465,-0.02705872803926468,-0.2062506377696991,-0.032158512622117996,-0.001065562479197979,-0.0021600360050797462,0.010404354892671108,-0.004580553621053696,-0.018738871440291405,0.006836488842964172,0.012025109492242336,-0.006518018431961536,-0.015243074856698513,-0.014231253415346146,-0.007296913769096136,-0.029546253383159637,-0.0028502221684902906,-0.002059105783700943,0.007829741574823856,-0.004269144497811794,-0.003506772220134735,0.0058273510076105595,0.038228943943977356,-0.012362856417894363,0.009196849539875984,0.0353967659175396,0.019691014662384987,0.043394770473241806,-0.02336663194000721,-0.02038705162703991,0.010309695266187191,0.010493269190192223,-0.0331498458981514,-0.00010303814633516595,4.527062628767453e-05,0.02266409434378147,0.0237970482558012,0.013503585942089558,0.033019356429576874,0.00894979014992714,0.026262132450938225,-0.0324513241648674,-0.0023145636077970266,-0.023432565852999687,0.025271620601415634,-0.054976947605609894,-0.04756287485361099,-0.0010525790276005864,-0.011116070672869682,-0.013173092156648636,0.013781415298581123,-0.04248414933681488,0.03725264593958855,0.02580961212515831,-0.012125491164624691,0.02321331948041916,-0.02967040054500103,0.006719142198562622,0.03681555762887001,0.012434516102075577,-0.020225614309310913,0.0033314889296889305,0.045238882303237915,0.0066699073649942875,-0.012408467009663582,0.00021850998746231198,0.019060451537370682,-0.0006506550125777721,0.01005454920232296,0.01281825453042984,0.00896813441067934,0.0012294386979192495,-0.019075196236371994,-0.010988892987370491,-0.05445272848010063,0.014546976424753666,-0.017002645879983902,-0.0025320511776953936,0.010421212762594223,0.005355092231184244,0.03140374645590782,-0.026313388720154762,-0.012697379104793072,0.02073219232261181,-0.030334170907735825,-0.002526636002585292,0.0777592658996582,-0.0059554907493293285,0.02468961291015148,-0.00607061805203557,-0.02240404114127159,0.006578207015991211,-0.08804464340209961,0.00022360611183103174,0.015509969554841518,0.0393218956887722,0.04558814689517021,-0.02600737102329731,-0.021574974060058594,0.017614254727959633,0.026419052854180336,-0.04863389581441879,-0.0016794393304735422,-0.016488557681441307,0.010242369025945663,-0.03042541816830635,0.05223585292696953,0.04129870608448982,0.01913716457784176,-0.041871797293424606,0.0030736117623746395,-0.01691257767379284,-0.013012166135013103,0.00914959516376257,-0.0024987095966935158,0.007190996780991554,-0.025641370564699173,-0.010286902077496052,-0.007655346300452948,0.013065553270280361,-0.034356068819761276,0.015048309229314327,-0.02974660135805607,-0.005650005768984556,-0.038438260555267334,-0.02338286116719246,0.011432373896241188,-0.026035446673631668,-0.03573842719197273,0.008249914273619652,-0.0048288218677043915,-0.009025415405631065,0.03204389289021492,-0.052015554159879684,-0.0005947394529357553,0.021108917891979218,-0.022609738633036613,0.007677216082811356,-0.002741162432357669,0.020198849961161613,-0.030973486602306366,0.022916922345757484,0.03197067603468895,0.0075868177227675915,0.021733008325099945,-0.022171737626194954,0.005577221978455782,0.06291462481021881,0.010051161050796509,0.005346635356545448,-0.0051164403557777405,0.0022031276021152735,-0.024477621540427208,0.020949041470885277,0.010851479135453701,0.038324132561683655,-0.006827448029071093,0.03552427887916565,-0.014253451488912106,0.015789812430739403,-0.008592965081334114,0.0009922918397933245,-0.02933741733431816,-0.006749261170625687,0.0038825541269034147,-0.0180740337818861,-0.019547542557120323,0.016794750466942787,0.01443388406187296,-0.01414264552295208,-0.0030532937962561846,-0.03292255103588104,-0.014052394777536392,-0.02198878861963749,-0.047747328877449036,0.0030371444299817085,0.021263878792524338,0.03646152466535568,-0.019917698577046394,0.0015535950660705566,-0.01240338385105133,0.011204435490071774,-0.004732639994472265,-0.005655153654515743,0.012495842762291431,-0.02762053720653057,0.0033291024155914783,-0.043057601898908615,-0.01731264591217041,0.009004561230540276,-0.010167465545237064,-0.017755435779690742,-0.006926957983523607,0.016514629125595093,-0.009980103000998497,-0.013384087942540646,-0.02238798327744007,0.011801606975495815,-0.02240007556974888,0.03967168927192688,0.005245546344667673,0.004888048395514488,-0.019460370764136314,-0.01258539967238903,-0.019087960943579674,-0.0111539326608181,-0.02416262961924076,0.0175357423722744,0.037282995879650116,-0.009718337096273899,0.004204864148050547,-0.05133983865380287,-0.01688362844288349,0.007046089507639408,0.030954411253333092,-0.04153471067547798,-0.045260824263095856,-0.008186589926481247,0.03154851496219635,-0.011969528160989285,0.0371565967798233,0.025364214554429054,0.006151722744107246,-0.00727783003821969,-0.009958743117749691,-0.009278294630348682,0.007596738636493683,0.01336624100804329,-0.021447397768497467,0.02054646424949169,-0.02611776813864708,-0.00033270398853346705,-0.03493153676390648,0.019608505070209503,0.0039603435434401035,-0.06740758568048477,-0.006607137154787779,-0.015215092338621616,0.0280825924128294,-0.01991902105510235,0.03201878070831299,-0.04012446105480194,-0.3158552050590515,0.01535746268928051,-0.036682210862636566,-0.010678562335669994,-0.019429007545113564,0.021722931414842606,0.0052073015831410885,-0.01830298826098442,0.011535142548382282],\n \"object\":\"embedding\"}],\n \"usage\":{\"num_images\":1,\"prompt_tokens\":77,\"total_tokens\":77},\n \"model\":\"nvidia/nvclip\"\n}\n"])</script><script>self.__next_f.push([1,"c6:T4d7,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/nvclip:2.0.0\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X POST 'http://0.0.0.0:8000/v1/embeddings' \\\n -H \"Content-Type: application/json\" \\\n -d '{\n \"input\": [\"The quick brown fox jumped over the lazy dog\",\n \"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAIAAACQd1PeAAAAEElEQVR4nGK6HcwNCAAA//8DTgE8HuxwEQAAAABJRU5ErkJggg==\"\n ],\n \"model\": \"nvidia/nvclip-vit-h-14\",\n \"encoding_format\": \"float\"\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/nvclip/latest/introduction.html).\nc7:Ted2,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/ingestion/table-extraction/latest/support-matrix.html#supported-hardware))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod o+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/nvclip:2.0.0\n```\n\n\n## Step 3. Test the NIM\n\nYou can now make a local API call by opening another Distro instance and using this curl command:\n\n```bash\ncurl -X POST 'http://0.0.0.0:8000/v1/embeddings' \\\n -H \"Content-Type: application/json\" \\\n -d '{\n \"input\": [\"The quick brown fox jumped over the lazy dog\",\n \"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAIAAACQd1PeAAAAEElEQVR4nGK6HcwNCAAA//8DTgE8HuxwEQAAAABJRU5ErkJggg==\"\n ],\n \"model\": \"nvidia/nvclip-vit-h-14\",\n \"encoding_format\": \"float\"\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/nvclip/latest/getting-started.html).\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\n"])</script><script>self.__next_f.push([1,"c8:Tdef,"])</script><script>self.__next_f.push([1,"## Model Overview\n\n### Description\n\nPaddleOCR is an ultra-lightweight Optical Character Recognition (OCR) system developed by Baidu. It supports a variety of cutting-edge OCR algorithms and provides value at every stage of the AI pipeline, including data generation, model training, and inference.\n\nThis model is ready for commercial use.\n\n## Third-Party Community Consideration\nThis model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA [PaddleOCR Toolkit](https://github.com/PaddlePaddle/PaddleOCR).\n\n### Terms of use\nPaddleOCR is licensed under [Apache-2](https://www.apache.org/licenses/LICENSE-2.0).\n**You are responsible for ensuring that your use of models complies with all applicable laws.**\n\n### References\n[Github](https://github.com/PaddlePaddle/PaddleOCR/blob/main/README_en.md)\n[Arxiv](https://arxiv.org/abs/2206.03001)\n\n\n## Model Architecture\n**Architecture Type for Text Detector:** CNN \u003cbr\u003e\n**Network Architecture for Text Detector:** LK-PAN\n\n**Architecture Type for Text Recognition:** Hybrid Transformer CNN \u003cbr\u003e\n**Network Architecture for Text Recognition:** SVTR-LCNet (NRTR Head and CTCLoss head) \u003cbr\u003e\n\n## Input\n**Input Type(s):** Image \u003cbr\u003e\n**Input Format(s):** Red, Green, Blue (RGB) \u003cbr\u003e\n**Input Parameters:** Two Dimensional (2D) \u003cbr\u003e\n**Other Properties Related to Input:** nd array, or batch of nd arrays are passed in with shape [Batch, Channel, Width, Height]. PaddleOCR does some internal thresholding, but none was implemented from our side. \u003cbr\u003e\n\n## Output\n**Output Type(s):** Text \u003cbr\u003e\n**Output Format:** String \u003cbr\u003e\n**Output Parameters:** 1D \u003cbr\u003e\n**Other Properties Related to Output:** Batch of text strings. \u003cbr\u003e\n\n**Supported Hardware Microarchitecture Compatibility:** NVIDIA Ampere, NVIDIA Hopper, NVIDIA Lovelace\u003cbr\u003e\n\n## Supported Operating System(s):\n* Linux \u003cbr\u003e\n\n## Model Version(s):\n* baidu/paddleocr \u003cbr\u003e\n\n## Training Dataset:\n\n**Link:** \u003cbr\u003e\n\nText detection datasets include LSVT (Sun et al. 2019), RCTW-17 (Shiet al. 2017), MTWI 2018 (He and Yang 2018), CASIA-10K (He et al. 2018), SROIE (Huang et al. 2019), MLT 2019 (Nayef et al. 2019), BDI (Karatzas et al. 2011), MSRATD500 (Yao et al. 2012) and CCPD 2019 (Xu et al. 2018).\n\nThese are two of the datasets (among others) which are used for text recognition:\n[OpenImages](https://github.com/openimages/dataset) \u003cbr\u003e\n[InvoiceDatasets](https://github.com/FuxiJia/InvoiceDatasets)\n\n**Data Collection Method by dataset:** Unknown \u003cbr\u003e\n**Labeling Method by dataset** Unknown \u003cbr\u003e\n\nText Detection: 127k training images (68K real scene images from Baidu image search and public datasets and 59K synthetic images)\n\nText Recognition: 18.5M training images (7M real scene images from Baidu image search and public datasets and 11.5M synthetic images)\n\n## Inference:\n**Engine:** Tensor(RT) \u003cbr\u003e\n**Test Hardware:** Tested on all supported hardware listed in compatibility section \u003cbr\u003e\n\n## Ethical Considerations:\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.\n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"c9:T5791e,"])</script><script>self.__next_f.push([1,"{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,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\"\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"ca:T4cb0,"])</script><script>self.__next_f.push([1,"{\n \"data\": [\n {\n \"index\": 0,\n \"text_detections\": [\n {\n \"text_prediction\": {\n \"text\": \"Trunch Parish Council\",\n \"confidence\": 0.9192708134651184\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.33125000000000004,\n \"y\": 0.02227466504263094\n },\n {\n \"x\": 0.5531250000000001,\n \"y\": 0.01641291108404385\n },\n {\n \"x\": 0.5531250000000001,\n \"y\": 0.04103227771010962\n },\n {\n \"x\": 0.33125000000000004,\n \"y\": 0.04689403166869672\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BANK RECONCILIATION AS AT31STOCTOBER 2019\",\n \"confidence\": 0.9528748989105225\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.18958333333333335,\n \"y\": 0.0984774665042631\n },\n {\n \"x\": 0.7010416666666667,\n \"y\": 0.0855816077953715\n },\n {\n \"x\": 0.7010416666666667,\n \"y\": 0.11020097442143728\n },\n {\n \"x\": 0.18958333333333335,\n \"y\": 0.12309683313032889\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Account:\",\n \"confidence\": 0.940185546875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.011458333333333334,\n \"y\": 0.17585261875761266\n },\n {\n \"x\": 0.10625000000000001,\n \"y\": 0.17350791717417785\n },\n {\n \"x\": 0.10625000000000001,\n \"y\": 0.20164433617539587\n },\n {\n \"x\": 0.011458333333333334,\n \"y\": 0.20398903775883073\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"14,389.43\",\n \"confidence\": 0.9986979365348816\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.2286084043848965\n },\n {\n \"x\": 0.8822916666666667,\n \"y\": 0.22626370280146166\n },\n {\n \"x\": 0.8822916666666667,\n \"y\": 0.2555724725943971\n },\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.257917174177832\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BANK STATEMENT BALANCE 30TH SEPTEMBER 2019\",\n \"confidence\": 0.945068359375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.2508830694275274\n },\n {\n \"x\": 0.5416666666666667,\n \"y\": 0.235642509135201\n },\n {\n \"x\": 0.5416666666666667,\n \"y\": 0.26026187576126675\n },\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.2755024360535932\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9986327886581421\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8083333333333335,\n \"y\": 0.26377892813641907\n },\n {\n \"x\": 0.8833333333333334,\n \"y\": 0.2614342265529842\n },\n {\n \"x\": 0.884375,\n \"y\": 0.29074299634591966\n },\n {\n \"x\": 0.8093750000000001,\n \"y\": 0.29308769792935446\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"PREVIOUS OUTSTANDING CHEQUES\",\n \"confidence\": 0.9708949327468872\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.28605359317904994\n },\n {\n \"x\": 0.3791666666666667,\n \"y\": 0.27667478684531066\n },\n {\n \"x\": 0.3791666666666667,\n \"y\": 0.3012941534713764\n },\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.3106729598051157\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"14,305.83\",\n \"confidence\": 0.9437391757965088\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.300121802679659\n },\n {\n \"x\": 0.884375,\n \"y\": 0.29777710109622413\n },\n {\n \"x\": 0.884375,\n \"y\": 0.3259135200974422\n },\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.328258221680877\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"CASH BOOK BALANCE31STOCTOBER 2019\",\n \"confidence\": 0.9460670948028564\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.3212241169305725\n },\n {\n \"x\": 0.44895833333333335,\n \"y\": 0.3095006090133983\n },\n {\n \"x\": 0.44895833333333335,\n \"y\": 0.33411997563946405\n },\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.34584348355663824\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"ADD CHEQUES OUTSTANDING:\",\n \"confidence\": 0.9514973759651184\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.35756699147381243\n },\n {\n \"x\": 0.3291666666666667,\n \"y\": 0.34936053593179056\n },\n {\n \"x\": 0.3291666666666667,\n \"y\": 0.3728075517661389\n },\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.3810140073081609\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.98681640625\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.89375,\n \"y\": 0.3692904993909867\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.3692904993909867\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.38804811205846534\n },\n {\n \"x\": 0.89375,\n \"y\": 0.38804811205846534\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"101719\",\n \"confidence\": 0.999755859375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.6229166666666667,\n \"y\": 0.41032277710109627\n },\n {\n \"x\": 0.7041666666666667,\n \"y\": 0.40797807551766146\n },\n {\n \"x\": 0.7052083333333334,\n \"y\": 0.43845919610231426\n },\n {\n \"x\": 0.6239583333333334,\n \"y\": 0.4408038976857491\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9019531011581421\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8125000000000001,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.8958333333333335,\n \"y\": 0.4044610231425092\n },\n {\n \"x\": 0.8968750000000001,\n \"y\": 0.43376979293544465\n },\n {\n \"x\": 0.8135416666666667,\n \"y\": 0.4361144945188794\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.94482421875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.4255633373934227\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4255633373934227\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.9365234375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4396315468940317\n },\n {\n \"x\": 0.909375,\n \"y\": 0.4396315468940317\n },\n {\n \"x\": 0.909375,\n \"y\": 0.4642509135200975\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4642509135200975\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.98095703125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.47714677222898905\n },\n {\n \"x\": 0.909375,\n \"y\": 0.47714677222898905\n },\n {\n \"x\": 0.909375,\n \"y\": 0.5005937880633374\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.5005937880633374\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.982421875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.5146619975639465\n },\n {\n \"x\": 0.9083333333333334,\n \"y\": 0.5146619975639465\n },\n {\n \"x\": 0.9083333333333334,\n \"y\": 0.5322472594397076\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.5322472594397076\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9984375238418579\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8177083333333335,\n \"y\": 0.6940316686967114\n },\n {\n \"x\": 0.89375,\n \"y\": 0.6916869671132765\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.7209957369062119\n },\n {\n \"x\": 0.8187500000000001,\n \"y\": 0.7233404384896469\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"OUTSTANDING CHEQUES\",\n \"confidence\": 0.9662829041481018\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.025,\n \"y\": 0.7151339829476249\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.7092722289890377\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.7327192448233861\n },\n {\n \"x\": 0.025,\n \"y\": 0.7385809987819734\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"9,148.00\",\n \"confidence\": 0.981689453125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7885416666666667,\n \"y\": 0.7678897685749088\n },\n {\n \"x\": 0.89375,\n \"y\": 0.7643727161997564\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.793681485992692\n },\n {\n \"x\": 0.7895833333333334,\n \"y\": 0.7971985383678443\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"RECEIPTS\",\n \"confidence\": 0.995849609375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.02291666666666667,\n \"y\": 0.7843026796589525\n },\n {\n \"x\": 0.12187500000000001,\n \"y\": 0.7807856272838004\n },\n {\n \"x\": 0.12291666666666669,\n \"y\": 0.8089220462850184\n },\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8124390986601706\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"4,309.94\",\n \"confidence\": 0.96173095703125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7895833333333334,\n \"y\": 0.8054049939098661\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.8018879415347138\n },\n {\n \"x\": 0.89375,\n \"y\": 0.8276796589524971\n },\n {\n \"x\": 0.790625,\n \"y\": 0.8311967113276493\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"PAYMENTS\",\n \"confidence\": 0.99652099609375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8206455542021925\n },\n {\n \"x\": 0.14166666666666666,\n \"y\": 0.8171285018270403\n },\n {\n \"x\": 0.14166666666666666,\n \"y\": 0.8452649208282583\n },\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8487819732034105\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"19,227.49\",\n \"confidence\": 0.9561631679534912\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7781250000000001,\n \"y\": 0.8405755176613886\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.8394031668696712\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.8651948842874544\n },\n {\n \"x\": 0.7781250000000001,\n \"y\": 0.8663672350791718\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BALANCE 31STOCTOBER2019\",\n \"confidence\": 0.9579653739929199\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.028125,\n \"y\": 0.8569884287454325\n },\n {\n \"x\": 0.3354166666666667,\n \"y\": 0.8511266747868454\n },\n {\n \"x\": 0.3354166666666667,\n \"y\": 0.8757460414129111\n },\n {\n \"x\": 0.028125,\n \"y\": 0.8816077953714982\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"19,227.49*\",\n \"confidence\": 0.9873046875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7791666666666667,\n \"y\": 0.9109165651644338\n },\n {\n \"x\": 0.9156250000000001,\n \"y\": 0.9073995127892814\n },\n {\n \"x\": 0.9156250000000001,\n \"y\": 0.9367082825822168\n },\n {\n \"x\": 0.7791666666666667,\n \"y\": 0.9402253349573692\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BALANCE AS PER BANK STATEMENT\",\n \"confidence\": 0.9529061317443848\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.928501827040195\n },\n {\n \"x\": 0.39062500000000006,\n \"y\": 0.9214677222898905\n },\n {\n \"x\": 0.39062500000000006,\n \"y\": 0.9460870889159563\n },\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9531211936662607\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"0.00\",\n \"confidence\": 0.9998779296875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8364583333333334,\n \"y\": 0.9449147381242389\n },\n {\n \"x\": 0.8989583333333334,\n \"y\": 0.9449147381242389\n },\n {\n \"x\": 0.8989583333333334,\n \"y\": 0.9753958587088917\n },\n {\n \"x\": 0.8364583333333334,\n \"y\": 0.9753958587088917\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"DIFFERENCE\",\n \"confidence\": 0.9974609613418579\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9636723507917175\n },\n {\n \"x\": 0.15520833333333336,\n \"y\": 0.9613276492082827\n },\n {\n \"x\": 0.15520833333333336,\n \"y\": 0.9859470158343485\n },\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9882917174177832\n }\n ]\n }\n }\n ]\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"cb:T23092,"])</script><script>self.__next_f.push([1,"{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,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\"\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"cc:T2001,"])</script><script>self.__next_f.push([1,"{\n \"data\": [\n {\n \"index\": 0,\n \"text_detections\": [\n {\n \"text_prediction\": {\n \"text\": \"July 2019\",\n \"confidence\": 0.9659830927848816\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.453125,\n \"y\": 0.07253333333333332\n },\n {\n \"x\": 0.5520833333333333,\n \"y\": 0.07253333333333332\n },\n {\n \"x\": 0.5520833333333333,\n \"y\": 0.18133333333333335\n },\n {\n \"x\": 0.453125,\n \"y\": 0.18133333333333335\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Date/Month\",\n \"confidence\": 0.995068371295929\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.23573333333333335\n },\n {\n \"x\": 0.14479166666666665,\n \"y\": 0.23573333333333335\n },\n {\n \"x\": 0.14479166666666665,\n \"y\": 0.3264\n },\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.3264\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Events/ Activities/Holidays\",\n \"confidence\": 0.9811559319496155\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.30208333333333337,\n \"y\": 0.23573333333333335\n },\n {\n \"x\": 0.5791666666666666,\n \"y\": 0.2538666666666667\n },\n {\n \"x\": 0.5791666666666666,\n \"y\": 0.34906666666666664\n },\n {\n \"x\": 0.30208333333333337,\n \"y\": 0.33093333333333336\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"1-31July2019\",\n \"confidence\": 0.94921875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.34453333333333336\n },\n {\n \"x\": 0.19687500000000002,\n \"y\": 0.3581333333333333\n },\n {\n \"x\": 0.19687500000000002,\n \"y\": 0.45333333333333337\n },\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.43973333333333336\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Mid Semester Vacation for the Teaching Staffof the College\",\n \"confidence\": 0.958369791507721\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.30208333333333337,\n \"y\": 0.34906666666666664\n },\n {\n \"x\": 0.8770833333333333,\n \"y\": 0.3808\n },\n {\n \"x\": 0.8770833333333333,\n \"y\": 0.48506666666666665\n },\n {\n \"x\": 0.30208333333333337,\n \"y\": 0.45333333333333337\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"10h-16July2019\",\n \"confidence\": 0.958740234375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.4578666666666667\n },\n {\n \"x\": 0.21770833333333334,\n \"y\": 0.476\n },\n {\n \"x\": 0.21770833333333334,\n \"y\": 0.5848000000000001\n },\n {\n \"x\": 0.017708333333333333,\n \"y\": 0.5666666666666667\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Declaration of the B.A./ B.Sc./ B.Com. Even Semester Examination\",\n \"confidence\": 0.9426651000976562\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.3010416666666667,\n \"y\": 0.48053333333333337\n },\n {\n \"x\": 0.9833333333333334,\n \"y\": 0.5077333333333334\n },\n {\n \"x\": 0.9833333333333334,\n \"y\": 0.5983999999999999\n },\n {\n \"x\": 0.3010416666666667,\n \"y\": 0.5712\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Results.\",\n \"confidence\": 0.97442626953125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.30000000000000004,\n \"y\": 0.5802666666666666\n },\n {\n \"x\": 0.3770833333333333,\n \"y\": 0.5893333333333334\n },\n {\n \"x\": 0.37604166666666666,\n \"y\": 0.7026666666666668\n },\n {\n \"x\": 0.2989583333333333,\n \"y\": 0.6936\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Total No.of working days\",\n \"confidence\": 0.9542643427848816\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.016666666666666666,\n \"y\": 0.7072\n },\n {\n \"x\": 0.2791666666666667,\n \"y\": 0.7162666666666666\n },\n {\n \"x\": 0.2791666666666667,\n \"y\": 0.8069333333333334\n },\n {\n \"x\": 0.016666666666666666,\n \"y\": 0.7978666666666667\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"27\",\n \"confidence\": 0.9990234375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.29791666666666666,\n \"y\": 0.7072\n },\n {\n \"x\": 0.32916666666666666,\n \"y\": 0.7072\n },\n {\n \"x\": 0.32916666666666666,\n \"y\": 0.8205333333333333\n },\n {\n \"x\": 0.29791666666666666,\n \"y\": 0.8205333333333333\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Total No.of teaching days\",\n \"confidence\": 0.96728515625\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.015625,\n \"y\": 0.8250666666666667\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.8386666666666668\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.9293333333333333\n },\n {\n \"x\": 0.015625,\n \"y\": 0.9157333333333334\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"00\",\n \"confidence\": 0.973876953125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.30000000000000004,\n \"y\": 0.8296\n },\n {\n \"x\": 0.32708333333333334,\n \"y\": 0.8296\n },\n {\n \"x\": 0.32708333333333334,\n \"y\": 0.9293333333333333\n },\n {\n \"x\": 0.30000000000000004,\n \"y\": 0.9293333333333333\n }\n ]\n }\n }\n ]\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"cd:T4cb0,"])</script><script>self.__next_f.push([1,"{\n \"data\": [\n {\n \"index\": 0,\n \"text_detections\": [\n {\n \"text_prediction\": {\n \"text\": \"Trunch Parish Council\",\n \"confidence\": 0.9192708134651184\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.33125000000000004,\n \"y\": 0.02227466504263094\n },\n {\n \"x\": 0.5531250000000001,\n \"y\": 0.01641291108404385\n },\n {\n \"x\": 0.5531250000000001,\n \"y\": 0.04103227771010962\n },\n {\n \"x\": 0.33125000000000004,\n \"y\": 0.04689403166869672\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BANK RECONCILIATION AS AT31STOCTOBER 2019\",\n \"confidence\": 0.9528748989105225\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.18958333333333335,\n \"y\": 0.0984774665042631\n },\n {\n \"x\": 0.7010416666666667,\n \"y\": 0.0855816077953715\n },\n {\n \"x\": 0.7010416666666667,\n \"y\": 0.11020097442143728\n },\n {\n \"x\": 0.18958333333333335,\n \"y\": 0.12309683313032889\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"Account:\",\n \"confidence\": 0.940185546875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.011458333333333334,\n \"y\": 0.17585261875761266\n },\n {\n \"x\": 0.10625000000000001,\n \"y\": 0.17350791717417785\n },\n {\n \"x\": 0.10625000000000001,\n \"y\": 0.20164433617539587\n },\n {\n \"x\": 0.011458333333333334,\n \"y\": 0.20398903775883073\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"14,389.43\",\n \"confidence\": 0.9986979365348816\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.2286084043848965\n },\n {\n \"x\": 0.8822916666666667,\n \"y\": 0.22626370280146166\n },\n {\n \"x\": 0.8822916666666667,\n \"y\": 0.2555724725943971\n },\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.257917174177832\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BANK STATEMENT BALANCE 30TH SEPTEMBER 2019\",\n \"confidence\": 0.945068359375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.2508830694275274\n },\n {\n \"x\": 0.5416666666666667,\n \"y\": 0.235642509135201\n },\n {\n \"x\": 0.5416666666666667,\n \"y\": 0.26026187576126675\n },\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.2755024360535932\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9986327886581421\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8083333333333335,\n \"y\": 0.26377892813641907\n },\n {\n \"x\": 0.8833333333333334,\n \"y\": 0.2614342265529842\n },\n {\n \"x\": 0.884375,\n \"y\": 0.29074299634591966\n },\n {\n \"x\": 0.8093750000000001,\n \"y\": 0.29308769792935446\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"PREVIOUS OUTSTANDING CHEQUES\",\n \"confidence\": 0.9708949327468872\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.28605359317904994\n },\n {\n \"x\": 0.3791666666666667,\n \"y\": 0.27667478684531066\n },\n {\n \"x\": 0.3791666666666667,\n \"y\": 0.3012941534713764\n },\n {\n \"x\": 0.015625000000000003,\n \"y\": 0.3106729598051157\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"14,305.83\",\n \"confidence\": 0.9437391757965088\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.300121802679659\n },\n {\n \"x\": 0.884375,\n \"y\": 0.29777710109622413\n },\n {\n \"x\": 0.884375,\n \"y\": 0.3259135200974422\n },\n {\n \"x\": 0.7645833333333334,\n \"y\": 0.328258221680877\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"CASH BOOK BALANCE31STOCTOBER 2019\",\n \"confidence\": 0.9460670948028564\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.3212241169305725\n },\n {\n \"x\": 0.44895833333333335,\n \"y\": 0.3095006090133983\n },\n {\n \"x\": 0.44895833333333335,\n \"y\": 0.33411997563946405\n },\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.34584348355663824\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"ADD CHEQUES OUTSTANDING:\",\n \"confidence\": 0.9514973759651184\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.35756699147381243\n },\n {\n \"x\": 0.3291666666666667,\n \"y\": 0.34936053593179056\n },\n {\n \"x\": 0.3291666666666667,\n \"y\": 0.3728075517661389\n },\n {\n \"x\": 0.01666666666666667,\n \"y\": 0.3810140073081609\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.98681640625\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.89375,\n \"y\": 0.3692904993909867\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.3692904993909867\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.38804811205846534\n },\n {\n \"x\": 0.89375,\n \"y\": 0.38804811205846534\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"101719\",\n \"confidence\": 0.999755859375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.6229166666666667,\n \"y\": 0.41032277710109627\n },\n {\n \"x\": 0.7041666666666667,\n \"y\": 0.40797807551766146\n },\n {\n \"x\": 0.7052083333333334,\n \"y\": 0.43845919610231426\n },\n {\n \"x\": 0.6239583333333334,\n \"y\": 0.4408038976857491\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9019531011581421\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8125000000000001,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.8958333333333335,\n \"y\": 0.4044610231425092\n },\n {\n \"x\": 0.8968750000000001,\n \"y\": 0.43376979293544465\n },\n {\n \"x\": 0.8135416666666667,\n \"y\": 0.4361144945188794\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.94482421875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.406805724725944\n },\n {\n \"x\": 0.9062500000000001,\n \"y\": 0.4255633373934227\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4255633373934227\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.9365234375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4396315468940317\n },\n {\n \"x\": 0.909375,\n \"y\": 0.4396315468940317\n },\n {\n \"x\": 0.909375,\n \"y\": 0.4642509135200975\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.4642509135200975\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.98095703125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.47714677222898905\n },\n {\n \"x\": 0.909375,\n \"y\": 0.47714677222898905\n },\n {\n \"x\": 0.909375,\n \"y\": 0.5005937880633374\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.5005937880633374\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"*\",\n \"confidence\": 0.982421875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.5146619975639465\n },\n {\n \"x\": 0.9083333333333334,\n \"y\": 0.5146619975639465\n },\n {\n \"x\": 0.9083333333333334,\n \"y\": 0.5322472594397076\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.5322472594397076\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"83.60\",\n \"confidence\": 0.9984375238418579\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8177083333333335,\n \"y\": 0.6940316686967114\n },\n {\n \"x\": 0.89375,\n \"y\": 0.6916869671132765\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.7209957369062119\n },\n {\n \"x\": 0.8187500000000001,\n \"y\": 0.7233404384896469\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"OUTSTANDING CHEQUES\",\n \"confidence\": 0.9662829041481018\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.025,\n \"y\": 0.7151339829476249\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.7092722289890377\n },\n {\n \"x\": 0.28020833333333334,\n \"y\": 0.7327192448233861\n },\n {\n \"x\": 0.025,\n \"y\": 0.7385809987819734\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"9,148.00\",\n \"confidence\": 0.981689453125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7885416666666667,\n \"y\": 0.7678897685749088\n },\n {\n \"x\": 0.89375,\n \"y\": 0.7643727161997564\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.793681485992692\n },\n {\n \"x\": 0.7895833333333334,\n \"y\": 0.7971985383678443\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"RECEIPTS\",\n \"confidence\": 0.995849609375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.02291666666666667,\n \"y\": 0.7843026796589525\n },\n {\n \"x\": 0.12187500000000001,\n \"y\": 0.7807856272838004\n },\n {\n \"x\": 0.12291666666666669,\n \"y\": 0.8089220462850184\n },\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8124390986601706\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"4,309.94\",\n \"confidence\": 0.96173095703125\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7895833333333334,\n \"y\": 0.8054049939098661\n },\n {\n \"x\": 0.8927083333333334,\n \"y\": 0.8018879415347138\n },\n {\n \"x\": 0.89375,\n \"y\": 0.8276796589524971\n },\n {\n \"x\": 0.790625,\n \"y\": 0.8311967113276493\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"PAYMENTS\",\n \"confidence\": 0.99652099609375\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8206455542021925\n },\n {\n \"x\": 0.14166666666666666,\n \"y\": 0.8171285018270403\n },\n {\n \"x\": 0.14166666666666666,\n \"y\": 0.8452649208282583\n },\n {\n \"x\": 0.023958333333333335,\n \"y\": 0.8487819732034105\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"19,227.49\",\n \"confidence\": 0.9561631679534912\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7781250000000001,\n \"y\": 0.8405755176613886\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.8394031668696712\n },\n {\n \"x\": 0.8947916666666667,\n \"y\": 0.8651948842874544\n },\n {\n \"x\": 0.7781250000000001,\n \"y\": 0.8663672350791718\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BALANCE 31STOCTOBER2019\",\n \"confidence\": 0.9579653739929199\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.028125,\n \"y\": 0.8569884287454325\n },\n {\n \"x\": 0.3354166666666667,\n \"y\": 0.8511266747868454\n },\n {\n \"x\": 0.3354166666666667,\n \"y\": 0.8757460414129111\n },\n {\n \"x\": 0.028125,\n \"y\": 0.8816077953714982\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"19,227.49*\",\n \"confidence\": 0.9873046875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.7791666666666667,\n \"y\": 0.9109165651644338\n },\n {\n \"x\": 0.9156250000000001,\n \"y\": 0.9073995127892814\n },\n {\n \"x\": 0.9156250000000001,\n \"y\": 0.9367082825822168\n },\n {\n \"x\": 0.7791666666666667,\n \"y\": 0.9402253349573692\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"BALANCE AS PER BANK STATEMENT\",\n \"confidence\": 0.9529061317443848\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.928501827040195\n },\n {\n \"x\": 0.39062500000000006,\n \"y\": 0.9214677222898905\n },\n {\n \"x\": 0.39062500000000006,\n \"y\": 0.9460870889159563\n },\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9531211936662607\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"0.00\",\n \"confidence\": 0.9998779296875\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.8364583333333334,\n \"y\": 0.9449147381242389\n },\n {\n \"x\": 0.8989583333333334,\n \"y\": 0.9449147381242389\n },\n {\n \"x\": 0.8989583333333334,\n \"y\": 0.9753958587088917\n },\n {\n \"x\": 0.8364583333333334,\n \"y\": 0.9753958587088917\n }\n ]\n }\n },\n {\n \"text_prediction\": {\n \"text\": \"DIFFERENCE\",\n \"confidence\": 0.9974609613418579\n },\n \"bounding_box\": {\n \"points\": [\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9636723507917175\n },\n {\n \"x\": 0.15520833333333336,\n \"y\": 0.9613276492082827\n },\n {\n \"x\": 0.15520833333333336,\n \"y\": 0.9859470158343485\n },\n {\n \"x\": 0.030208333333333334,\n \"y\": 0.9882917174177832\n }\n ]\n }\n }\n ]\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"ce:T54a,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/baidu/paddleocr:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\nHOSTNAME=\"localhost\"\nSERVICE_PORT=8000\ncurl -X \"POST\" \\\n \"http://${HOSTNAME}:${SERVICE_PORT}/v1/infer\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n },\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n }\n ]\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/ingestion/table-extraction/latest/getting-started.html).cf:T1025,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/ingestion/table-extraction/latest/support-matrix.html#supported-hardware))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod o+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/baidu/paddleocr:1.1.0-rtx\n```\n\nThe first few (depending on number of GPUs) inference requests may take longer than subsequent ones. This is due to the model being loaded into memory and initialized for the first time.\n\n## Step 3. Test the NIM\n\nYou can now make a local API call by opening another Distro instance and using this curl command:\n\n```bash\nHOSTNAME=\"localhost\"\nSERVICE_PORT=8000\ncurl -X \"POST\" \\\n \"http://${HOSTNAME}:${SERVICE_PORT}/v1/infer\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n },\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n }\n ]\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/ingestion/table-extraction/latest/getting-started.html).\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;"])</script><script>self.__next_f.push([1,"d0:T359f,"])</script><script>self.__next_f.push([1,"## Model Overview\u003ca id=\"model-overview\"\u003e\u003c/a\u003e\n\n### Description\u003ca id=\"description\"\u003e\u003c/a\u003e\n\nYOLOX is an anchor-free version of YOLO (You Only Look Once) one-shot object detection model series, with a simpler design, better performance and less restrictive license. It’s from Megvii Technology. This model is a YOLOX-L version fine-tuned on 26,000 images from [Digital Corpora dataset](https://digitalcorpora.org/), with annotations from Azure AI Document Intelligence. The model is trained to detect tables, charts and titles in documents.\n\nThis model is ready for commercial use.\n\n### License/Terms of use\u003ca id=\"terms-of-use\"\u003e\u003c/a\u003e\n\n Use of this model is governed by the [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-ai-foundation-models-community-license-agreement/#:~:text=This%20license%20agreement%20(%E2%80%9CAgreement%E2%80%9D,algorithms%2C%20parameters%2C%20configuration%20files%2C)) and the [Apache 2.0 License](https://github.com/Megvii-BaseDetection/YOLOX/blob/main/LICENSE).\n\n### Model Architecture\u003ca id=\"model-architecture\"\u003e\u003c/a\u003e\n**Architecture Type:** Yolox\u003cbr\u003e\n**Network Architecture:** DarkNet53 Backbone + FPN Decoupled head (one 1x1 convolution + 2 parallel 3x3 convolutions (one for the classification and one for the bounding box prediction)\u003cbr\u003e\n\nYOLOX is a single-stage object detector that improves on Yolo-v3.\nThe model is fine-tuned to detect three classes of objects in documents: table, chart, title. Chart is defined as a bar chart, line chart or pie chart. Titles can be page titles, section titles, or table/chart titles.\n\n### Model Version(s)\u003ca id=\"model-versions\"\u003e\u003c/a\u003e\nShort name: YOLOX Document Structure Detection\n\n### Intended use\u003ca id=\"intended-use\"\u003e\u003c/a\u003e\nYOLOX Model is suitable for users who want to extract tables, text titles and charts from documents. It can be used for document analysis, document understanding, and document processing. The goal is to extract structural elements (tables and charts) from the page to allow vision models to be applied for information extraction.\n\n## Technical Details\u003ca id=\"technical-details\"\u003e\u003c/a\u003e\n\n### Input\u003ca id=\"input\"\u003e\u003c/a\u003e\n**Input Type(s):** Image \u003cbr\u003e\n**Input Format(s):** Red, Green, Blue (RGB) \u003cbr\u003e\n**Input Parameters:** Two Dimensional (2D) \u003cbr\u003e\n**Other Properties Related to Input:** Image size resized to (1024, 1024) \u003cbr\u003e\n\n### Output\u003ca id=\"output\"\u003e\u003c/a\u003e\n**Output Type(s):** Array \u003cbr\u003e\n**Output Format:** dict of dictionaries containing np.ndarray. Outer dictionary contains each sample (page). Inner dictionary contains list of dictionaries with bboxes, type and confidence for that page \u003cbr\u003e\n**Output Parameters:** n/a \u003cbr\u003e\n**Other Properties Related to Output:** Output contains Bounding box, detection confidence and object class (chart, table, title). Thresholds used for nms - conf_thresh = 0.01; iou_thresh = 0.5; max_per_img = 100; min_per_img = 0 \u003cbr\u003e\n\n### Software Integration\u003ca id=\"software-integration\"\u003e\u003c/a\u003e\n\n**Runtime:** NeMo Retriever YOLOX Structured Images NIM \u003cbr\u003e\n**Supported Hardware Microarchitecture Compatibility:** NVIDIA Ampere, NVIDIA Hopper, NVIDIA Lovelace\u003cbr\u003e\n\n## Supported Operating System(s):\n* Linux \u003cbr\u003e\n\n## Model Version(s):\n* nvidia/nv-yolox-structured-images-v1 \u003cbr\u003e\n\n## Training Dataset \u0026 Evaluation\u003ca id=\"training-dataset--evaluation\"\u003e\u003c/a\u003e\n\n\n### Training Dataset\u003ca id=\"training-dataset\"\u003e\u003c/a\u003e\n\n**Data Collection Method by dataset:** Automated\u003cbr\u003e\n**Labeling Method by dataset:** Automated \u003cbr\u003e\n\nPretraining: [COCO train2017](https://cocodataset.org/#download)\n\nFinetuning (by NVIDIA): 25,832 images from [Digital Corpora dataset](https://digitalcorpora.org/), with annotations from [Azure AI Document Intelligence](https://azure.microsoft.com/en-us/products/ai-services/ai-document-intelligence).\n\nNumber of bounding boxes per class: 30,099 tables, 34,369 titles and 8,363 charts. The layout model of Document Intelligence was used with `2024-02-29-preview` api version.\n\n### Evaluation Results\u003ca id=\"evaluation-results\"\u003e\u003c/a\u003e\n\nThe primary evaluation set is a cut of the azure labels and digital corpora images.\nNumber of bounding boxes per class: 1704 tables, 1906 titles and 367 charts. mAP was used as an evaluation metric.\n\n**Data Collection Method by dataset:** Automated \u003cbr\u003e\n**Labeling Method by dataset:** Automated, Human \u003cbr\u003e\n**Properties (Quantity, Dataset Descriptions, Sensor(s)):** We evaluated with azure labels from held out pages, as well as manual inspection on public PDFs and powerpoint slides.\n\n## Inference:\n**Engine:** TensorRT \u003cbr\u003e\n**Test Hardware:** See Support Matrix from NIM documentation. \u003cbr\u003e\n\n## Ethical Considerations\u003ca id=\"ethical-considerations\"\u003e\u003c/a\u003e\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ tab for the Explainability, Bias, Safety \u0026 Security, and Privacy subcards. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Model Card++\u003ca id=\"model-card\"\u003e\u003c/a\u003e\n\n### Bias\u003ca id=\"bias\"\u003e\u003c/a\u003e\n\n| | |\n| :--------------------------------------------------------------------------------------------------------------------------------------------------------------: | :----------: |\n| **Field** | **Response** |\n| Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing | None |\n| Measures taken to mitigate against unwanted bias | None |\n\n### Explainability\u003ca id=\"explainability\"\u003e\u003c/a\u003e\n\n| Field | Response |\n| :----------------------------: |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|\n| Intended Application \u0026 Domain: | Document Understanding |\n| Model Type: | Object Detection |\n| Intended User: | Enterprise developers who need to organise internal documentation |\n| Output: | Array of float numbers(with localisation information) |\n| Describe how the model works: | Model detects charts, tables and titles in an image. |\n| Verified to have met prescribed quality standards: | Yes\n| Performance Metrics: | Accuracy, Throughput, and Latency |\n| Potential Known Risks: | This model does not always guarantee to extract all entities in an image. |\n| Licensing \u0026 Terms of Use: | [NVIDIA AI Foundation Models Community License Agreement](https://developer.nvidia.com/downloads/nv-ai-foundation-models-license) and the [MIT License (MIT)](https://github.com/microsoft/unilm/blob/master/LICENSE). |\n| Technical Limitations | The model may not generalize to unknown document types not commonly found on the web. |\n\n### Privacy\u003ca id=\"privacy\"\u003e\u003c/a\u003e\n\n| Field | Response |\n| :---------------------------------------------------------------------------------------------------------------------------------------------: | :--------------------------------------------: |\n| Generatable or reverse engineerable personally-identifiable information (PII)? | Neither |\n| Was consent obtained for any personal data used? | Not Applicable |\n| Personal data used to create this model? | None |\n| How often is the dataset reviewed? | Before Every Release |\n| Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n| If personal data was collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable |\n| If personal data was collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable |\n| If personal data was collected for the development of this AI model, was it minimized to only what was required? | Not Applicable |\n| Is there provenance for all datasets used in training? | Yes |\n| Does data labeling (annotation, metadata) comply with privacy laws? | Yes |\n| Is data compliant with data subject requests for data correction or removal, if such a request was made? | No, not possible with externally-sourced data. |\n\n\n### Safety and Security\u003ca id=\"safety-and-security\"\u003e\u003c/a\u003e\n\n| Field | Response |\n| :------------------------------------------------: | :---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: |\n| Model Application(s): | Text Embedding for Retrieval |\n| Describe the physical safety impact (if present). | Not Applicable |\n| Use Case Restrictions: | Commercial Abide by [NVIDIA AI Foundation Models Community License Agreement](https://developer.nvidia.com/downloads/nv-ai-foundation-models-license). |\n| Model and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. |"])</script><script>self.__next_f.push([1,"d1:T22c93,"])</script><script>self.__next_f.push([1,"{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/jpeg;base64,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\"\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"d2:T56b,{\n \"data\": [\n {\n \"index\": 0,\n \"bounding_boxes\": {\n \"table\": [\n {\n \"x_min\": 0.6503,\n \"y_min\": 0.2161,\n \"x_max\": 0.7835,\n \"y_max\": 0.3236,\n \"confidence\": 0.9306\n },\n {\n \"x_min\": 0.2152,\n \"y_min\": 0.3378,\n \"x_max\": 0.3486,\n \"y_max\": 0.4352,\n \"confidence\": 0.9097\n },\n {\n \"x_min\": 0.5049,\n \"y_min\": 0.251,\n \"x_max\": 0.6366,\n \"y_max\": 0.3532,\n \"confidence\": 0.8976\n },\n {\n \"x_min\": 0.3603,\n \"y_min\": 0.2851,\n \"x_max\": 0.4909,\n \"y_max\": 0.3879,\n \"confidence\": 0.8886\n },\n {\n \"x_min\": 0.2267,\n \"y_min\": 0.8013,\n \"x_max\": 0.7689,\n \"y_max\": 0.8367,\n \"confidence\": 0.8753\n }\n ],\n \"chart\": [\n {\n \"x_min\": 0.2145,\n \"y_min\": 0.5498,\n \"x_max\": 0.7834,\n \"y_max\": 0.8422,\n \"confidence\": 0.9057\n }\n ],\n \"title\": [\n {\n \"x_min\": 0.2426,\n \"y_min\": 0.1387,\n \"x_max\": 0.7163,\n \"y_max\": 0.1891,\n \"confidence\": 0.7591\n }\n ]\n }\n }\n ]\n}\nd3:T236f3,"])</script><script>self.__next_f.push([1,"{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/jpeg;base64,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\"\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"d4:T40c,{\n \"data\": [\n {\n \"index\": 0,\n \"bounding_boxes\": {\n \"table\": [\n {\n \"x_min\": 0.2202,\n \"y_min\": 0.1853,\n \"x_max\": 0.9144,\n \"y_max\": 0.3686,\n \"confidence\": 0.9337\n }\n ],\n \"chart\": [\n {\n \"x_min\": 0.2695,\n \"y_min\": 0.4901,\n \"x_max\": 0.7304,\n \"y_max\": 0.734,\n \"confidence\": 0.7864\n }\n ],\n \"title\": [\n {\n \"x_min\": 0.2163,\n \"y_min\": 0.428,\n \"x_max\": 0.4164,\n \"y_max\": 0.4417,\n \"confidence\": 0.9405\n },\n {\n \"x_min\": 0.2154,\n \"y_min\": 0.758,\n \"x_max\": 0.3768,\n \"y_max\": 0.7728,\n \"confidence\": 0.8963\n },\n {\n \"x_min\": 0.3645,\n \"y_min\": 0.1562,\n \"x_max\": 0.632,\n \"y_max\": 0.1713,\n \"confidence\": 0.7569\n }\n ]\n }\n }\n ]\n}\nd5:T56b,{\n \"data\": [\n {\n \"index\": 0,\n \"bounding_boxes\": {\n \"table\": [\n {\n \"x_min\": 0.6503,\n \"y_min\": 0.2161,\n \"x_max\": 0.7835,\n \"y_max\": 0.3236,\n \"confidence\": 0.9306\n },\n {\n \"x_min\": 0.2152,\n \"y_min\": 0.3378,\n \"x_max\": 0.3486,\n \"y_max\": 0.4352,\n \"confidence\": 0.9097\n },\n {\n \"x_min\": 0.5049,\n \"y_min\": 0.251,\n \"x_max\": 0.6366,\n \"y_max\": 0.3532,\n \"confidence\": 0.8976\n },\n {\n \"x_min\": 0.3603,\n \"y_min\": 0.2851,\n \"x_max\": 0.4909,\n \"y_max\": 0.3879,\n \"confidence\": 0.8886\n },\n {\n \"x_min\": 0.2267,\n \"y_min\": 0.8013,\n \"x_max\": 0.7689,\n \"y_max\": 0.8367,\n \"confidence\": 0.8753\n }\n ],\n \"chart\": [\n {\n "])</script><script>self.__next_f.push([1," \"x_min\": 0.2145,\n \"y_min\": 0.5498,\n \"x_max\": 0.7834,\n \"y_max\": 0.8422,\n \"confidence\": 0.9057\n }\n ],\n \"title\": [\n {\n \"x_min\": 0.2426,\n \"y_min\": 0.1387,\n \"x_max\": 0.7163,\n \"y_max\": 0.1891,\n \"confidence\": 0.7591\n }\n ]\n }\n }\n ]\n}\nd6:T58c,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/nv-yolox-page-elements-v1:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\nHOSTNAME=\"localhost\"\nSERVICE_PORT=8000\ncurl -X \"POST\" \\\n \"http://${HOSTNAME}:${SERVICE_PORT}/v1/infer\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n },\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n }\n ]\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/ingestion/object-detection/latest/getting-started.html#get-started-with-nvidia-nim-for-object-detection).d7:T1063,"])</script><script>self.__next_f.push([1,"## Prerequisites\n\n* NVIDIA GeForce RTX 4080 or above (see [supported GPUs](https://docs.nvidia.com/nim/ingestion/object-detection/latest/support-matrix.html#nemo-retriever-yolox-page-elements-v1))\n* Install the latest [NVIDIA GPU Driver](https://www.nvidia.com/en-us/drivers/) on Windows (Version 570+)\n* Ensure virtualization is enabled in the system BIOS. In Windows, open Task Manager, select the Performance tab, and find Virtualization. If Disabled, see [here](https://support.microsoft.com/en-us/windows/enable-virtualization-on-windows-c5578302-6e43-4b4b-a449-8ced115f58e1) to enable.\n\n## Step 1. Open the Windows Subsystem for Linux 2 - WSL2 - Distro\n\n[Install WSL2](https://assets.ngc.nvidia.com/products/api-catalog/rtx/NIM_Prerequisites_Installer_03052025.zip). For additional instructions refer to the [documentation](https://docs.nvidia.com/nim/wsl2/latest/getting-started.html#installation).\n\nOnce installed, open the ``NVIDIA-Workbench`` WSL2 distro using the following command in the Windows terminal.\n\n```\nwsl -d NVIDIA-Workbench\n```\n\n## Step 2. Run the Container\n\n::generate-api-key\n\n\u003cp\u003e\u003c/p\u003e\n\n```bash\n$ podman login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\nchmod o+w \"$LOCAL_NIM_CACHE\"\npodman run -it --rm \\\n --device nvidia.com/gpu=all \\\n --shm-size=16GB \\\n -e NGC_API_KEY=$NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -e NIM_RELAX_MEM_CONSTRAINTS=1 \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/nvidia/nv-yolox-page-elements-v1:1.1.0-rtx\n```\n\nThe first few inference requests may take longer than subsequent ones. This is due to the model being loaded into memory and initialized for the first time.\n\n## Step 3. Test the NIM\n\nYou can now make a local API call by opening another Distro instance and using this curl command:\n\n```bash\nHOSTNAME=\"localhost\"\nSERVICE_PORT=8000\ncurl -X \"POST\" \\\n \"http://${HOSTNAME}:${SERVICE_PORT}/v1/infer\" \\\n -H 'accept: application/json' \\\n -H 'Content-Type: application/json' \\\n -d '{\n \"input\": [\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n },\n {\n \"type\": \"image_url\",\n \"url\": \"data:image/png;base64,\u003cBASE64_ENCODED_IMAGE\u003e\"\n }\n ]\n }'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/ingestion/object-detection/latest/getting-started.html#get-started-with-nvidia-nim-for-object-detection).\n\n\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;"])</script><script>self.__next_f.push([1,"d8:{\"name\":\"deepseek-r1-distill-llama-8b\",\"type\":\"model\"}\nd9:{\"name\":\"mistral-nemo-12b-instruct\",\"type\":\"model\"}\nda:{\"name\":\"llama-3_1-8b-instruct\",\"type\":\"model\"}\ndb:{\"name\":\"parakeet-ctc-0_6b-asr\",\"type\":\"model\"}\ndc:{\"name\":\"studiovoice\",\"type\":\"model\"}\ndd:{\"name\":\"llama-3_2-nv-embedqa-1b-v2\",\"type\":\"model\"}\nde:{\"name\":\"nvclip\",\"type\":\"model\"}\ndf:{\"name\":\"paddleocr\",\"type\":\"model\"}\ne0:{\"name\":\"nv-yolox-page-elements-v1\",\"type\":\"model\"}\n"])</script><script>self.__next_f.push([1,"39:[\"$\",\"$L3d\",null,{\"data\":[{\"endpoint\":{\"artifact\":{\"name\":\"deepseek-r1-distill-llama-8b\",\"displayName\":\"deepseek-r1-distill-llama-8b\",\"publisher\":\"deepseek-ai\",\"shortDescription\":\"Distilled version of Llama 3.1 8B using reasoning data generated by DeepSeek R1 for enhanced performance.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/deepseek-r1-distill-llama-8b.jpg\",\"labels\":[\"Distillation\",\"coding\",\"math\",\"reasoning\",\"run on rtx\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-03-17T17:48:42.953Z\",\"description\":\"$99\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:42:02.792Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"35245106-cdb6-4cee-ae1c-425bd80e468a\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for deepseek-ai/deepseek-r1-distill-llama-8b\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1-distill-llama-8b for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"MIT License\",\"url\":\"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B/blob/main/LICENSE\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletion\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionChunk\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Which number is larger, 9.11 or 9.8?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Which number is larger, 9.11 or 9.8?\\\"\\n }\\n ],\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The number 9.11 is...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"How many 'r's are in the word 'strawberry'?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"How many 'r's are in the word 'strawberry'?\\\"\\n }\\n ],\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"In the word strawberry, there are...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"$9a\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"$9b\",\"curl\":\"$9c\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1-distill-llama-8b\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"ChatCompletion\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/Choice\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Usage\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"},\"ChatCompletionChunk\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/ChoiceChunk\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"},\"ChatRequest\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"deepseek-ai/deepseek-r1-distill-llama-8b\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.6,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":4096,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"},\"Choice\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"ChoiceChunk\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"Message\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"user\",\"assistant\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"Usage\":{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:42:03.458Z\",\"nvcfFunctionId\":\"daa3e658-bf63-49bb-ac12-5c37edb8879f\",\"createdDate\":\"2025-03-17T17:48:43.209Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1-distill-llama-8b\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/research/terms/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Community Model License\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eMIT License\u003c/a\u003e; \u003ca href=\\\"https://www.llama.com/llama3_1/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.1 Community License Agreement\u003c/a\u003e. Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/deepseek-ai/containers/deepseek-r1-distill-llama-8b\"},\"playground\":{\"type\":\"chatWithThinking\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$9d\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$9e\"}]},\"artifactName\":\"deepseek-r1-distill-llama-8b\"},\"config\":{\"name\":\"deepseek-r1-distill-llama-8b\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"mistral-nemo-12b-instruct\",\"displayName\":\"mistral-nemo-12b-instruct\",\"publisher\":\"nv-mistralai\",\"shortDescription\":\"Most advanced language model for reasoning, code, multilingual tasks; runs on a single GPU.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/mistral-nemo-12b-instruct.jpg\",\"labels\":[\"Chat\",\"Code Generation\",\"Language Generation\",\"Text-to-text\",\"run on rtx\",\"Code Generation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-07-18T13:59:22.242Z\",\"description\":\"$9f\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T15:11:46.427Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"270991d0-1988-4b10-8e48-30a53510a113\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nv-mistralai/mistral-nemo-12b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nv-mistralai-mistral-nemo-12b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Apache 2.0\",\"url\":\"https://mistral.ai/terms-of-service/\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletion\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionChunk\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"Tell me about Dumbledore.\",\"requestJson\":\"$a0\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"What is the weather in Santa Clara, CA?\",\"requestJson\":\"$a1\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n base_url = \\\"https://integrate.api.nvidia.com/v1\\\",\\n api_key = \\\"$NVIDIA_API_KEY\\\"\\n)\\n\\ncompletion = client.chat.completions.create(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in completion:\\n if chunk.choices[0].delta.content is not None:\\n print(chunk.choices[0].delta.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nprint(completion.choices[0].message)\\n\u003c% } %\u003e\\n\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"import OpenAI from 'openai';\\n\\nconst openai = new OpenAI({\\n apiKey: '$NVIDIA_API_KEY',\\n baseURL: 'https://integrate.api.nvidia.com/v1',\\n})\\n\\nasync function main() {\\n const completion = await openai.chat.completions.create({\\n model: \\\"\u003c%- request.model %\u003e\\\",\\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature: \u003c%- request.temperature %\u003e,\\n top_p: \u003c%- request.top_p %\u003e,\\n max_tokens: \u003c%- request.max_tokens %\u003e,\\n stream: \u003c%- request.stream %\u003e,\\n })\\n \u003c% if (request.stream) { %\u003e\\n for await (const chunk of completion) {\\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\\n }\\n \u003c% } else { %\u003e\\n process.stdout.write(completion.choices[0]?.message?.content);\\n \u003c% } %\u003e\\n}\\n\\nmain();\",\"curl\":\"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature %\u003e, \\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e \\n }'\\n\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"nv-mistralai/mistral-nemo-12b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"ChatCompletion\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/Choice\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Usage\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"},\"ChatCompletionChunk\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/ChoiceChunk\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"},\"ChatRequest\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"nv-mistralai/mistral-nemo-12b-instruct\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":8192,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"},\"Choice\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"ChoiceChunk\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"Message\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"system\",\"context\",\"user\",\"assistant\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"type\":\"string\"}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"Usage\":{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T15:11:47.156Z\",\"nvcfFunctionId\":\"f8c05193-d2e2-4f0f-bb4d-7ad70070002b\",\"createdDate\":\"2024-07-18T13:59:22.640Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nv-mistralai-mistral-nemo-12b-instruct\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e. The use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e. ADDITIONAL INFORMATION: Apache 2.0.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/nv-mistralai/containers/mistral-nemo-12b-instruct\"},\"playground\":{\"type\":\"chatWithTools\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$a2\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$a3\"}]},\"artifactName\":\"mistral-nemo-12b-instruct\"},\"config\":{\"name\":\"mistral-nemo-12b-instruct\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"llama-3_1-8b-instruct\",\"displayName\":\"llama-3.1-8b-instruct\",\"publisher\":\"meta\",\"shortDescription\":\"Advanced state-of-the-art model with language understanding, superior reasoning, and text generation.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llama-3_1-8b-instruct.jpg\",\"labels\":[\"Chat\",\"Language Generation\",\"Run on RTX\",\"Text-to-Text\",\"Code Generation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-07-23T14:58:17.373Z\",\"description\":\"$a4\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:57:32.515Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"a1de03f9-a9b7-4f1b-8e72-47637b3dec2c\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for meta/llama-3.1-8b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/meta-llama-3_1-8b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Llama 3.1 License\",\"url\":\"https://github.com/meta-llama/llama-models/blob/main/License/Llama3.1.txt\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletion\"}},\"text/event-stream\":{\"schema\":{\"$ref\":\"#/components/schemas/ChatCompletionChunk\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"Tell me about Dumbledore.\",\"requestJson\":\"$a5\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"What is the weather in Santa Clara, CA?\",\"requestJson\":\"$a6\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"$a7\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n \\n\",\"node.js\":\"$a8\",\"curl\":\"$a9\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.1-8b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"ChatCompletion\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/Choice\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Usage\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"},\"ChatCompletionChunk\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"$ref\":\"#/components/schemas/ChoiceChunk\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"},\"ChatRequest\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"meta/llama-3.1-8b-instruct\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"tools\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/ChatCompletionToolsParam\"},\"type\":\"array\"},{\"type\":\"null\"}],\"title\":\"Tools\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"},\"Choice\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\",\"tool_calls\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"ChoiceChunk\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Message\"}],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\",\"tool_calls\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"Message\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"system\",\"context\",\"user\",\"assistant\",\"tool\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"tool_call_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Tool Call Id\",\"description\":\"The id of the tool call.\"},\"tool_calls\":{\"anyOf\":[{\"items\":{\"$ref\":\"#/components/schemas/ToolCall\"},\"type\":\"array\"},{\"type\":\"null\"}],\"title\":\"Tool Calls\",\"description\":\"The tool(s) called by the model.\"}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"ToolCall\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"},\"function\":{\"$ref\":\"#/components/schemas/FunctionCall\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ToolCall\"},\"FunctionCall\":{\"properties\":{\"name\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Name\"},\"arguments\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Arguments\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\",\"arguments\"],\"title\":\"FunctionCall\"},\"ChatCompletionToolsParam\":{\"properties\":{\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"},\"function\":{\"$ref\":\"#/components/schemas/FunctionDefinition\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ChatCompletionToolsParam\"},\"ChatCompletionNamedFunction\":{\"properties\":{\"name\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Name\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\"],\"title\":\"ChatCompletionNamedFunction\"},\"ChatCompletionNamedToolChoiceParam\":{\"properties\":{\"function\":{\"$ref\":\"#/components/schemas/ChatCompletionNamedFunction\"},\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ChatCompletionNamedToolChoiceParam\"},\"FunctionDefinition\":{\"properties\":{\"name\":{\"type\":\"string\",\"title\":\"Name\"},\"description\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Description\"},\"parameters\":{\"anyOf\":[{\"type\":\"object\"},{\"type\":\"null\"}],\"title\":\"Parameters\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\"],\"title\":\"FunctionDefinition\"},\"Usage\":{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:57:33.219Z\",\"nvcfFunctionId\":\"e62a4350-2218-4cf5-9262-112432d239f8\",\"createdDate\":\"2024-07-23T14:58:17.697Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/meta-llama-3_1-8b\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e AI Foundation Models Community License Agreement \u003c/a\u003e. ADDITIONAL INFORMATION: Llama 3.1 Community License Agreement, Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/meta/containers/llama-3.1-8b-instruct\"},\"playground\":{\"type\":\"chatWithTools\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$aa\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$ab\"}]},\"artifactName\":\"llama-3_1-8b-instruct\"},\"config\":{\"name\":\"llama-3_1-8b-instruct\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"parakeet-ctc-0_6b-asr\",\"displayName\":\"parakeet-ctc-0.6b-asr\",\"publisher\":\"nvidia\",\"shortDescription\":\"State-of-the-art accuracy and speed for English transcriptions.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/parakeet-ctc-0_6b-asr.jpg\",\"labels\":[\"ASR\",\"Batch\",\"English\",\"Fast\",\"NVIDIA NIM\",\"Run on RTX\",\"Streaming\",\"Speech-to-Text\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"Field | Response\\n:---------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\\nWhat is the language balance of the model validation data? | English: 100%\\nWhat is the geographic origin language balance of the model validation data? | United States: 80%, United Kingdom: 10%, Others: 10%\\nWhat is the accent balance of the model validation data? | American English: 80%, British English: 10%, Others: 10%\\nParticipation considerations from adversely impacted groups ([protected classes](https://www.senate.ca.gov/content/protected-classes)) in model design and testing: | Age, Gender, Linguistic Background\\nMeasures taken to mitigate against unwanted bias: | Used custom dataset to validate model performance across gender, age, and linguistic demographics\",\"canGuestDownload\":true,\"createdDate\":\"2024-08-06T06:27:13.040Z\",\"description\":\"$ac\",\"explainability\":\"$ad\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$ae\",\"safetyAndSecurity\":\"Field | Response\\n:---------------------------------------------------|:----------------------------------\\nModel Application(s): | Speech Transcription\\nDescribe the life-critical impacts (if present). | Not Applicable\\nUse Case Restriction(s): | Abide by https://developer.nvidia.com/riva/ga/license\\nDescribe access restrictions (if any): | The Principle of Least Privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training and dataset license constraints adhered to.\",\"updatedDate\":\"2025-03-24T23:59:02.186Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"2d149c7b-c46f-45e1-930b-fcd68059c0ea\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:59:03.026Z\",\"nvcfFunctionId\":\"d8dd4e9b-fbf5-4fb0-9dba-8cf436c8d965\",\"createdDate\":\"2024-08-06T06:27:13.383Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.nvidia.com/nim/riva/asr/latest/protos.html\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: Your use of this API is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Run Anywhere - Notify Me\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/parakeet-0-6b-ctc-en-us\"},\"usage\":\"$af\",\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$b0\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$b1\"}]},\"artifactName\":\"parakeet-ctc-0_6b-asr\"},\"config\":{\"name\":\"parakeet-ctc-0_6b-asr\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"studiovoice\",\"displayName\":\"studiovoice\",\"publisher\":\"nvidia\",\"shortDescription\":\"Enhance speech by correcting common audio degradations to create studio quality speech output.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/studiovoice.jpg\",\"labels\":[\"Digital Human\",\"Nvidia Maxine\",\"Run on RTX\",\"Speech Enhancement\",\"Speech-to-speech\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"| Field | Response |\\n|-------|----------|\\n| Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | Age (18+), Gender |\\n| Measures taken to mitigate against unwanted bias: | Evaluated using internal, proprietary data mix to achieve similar key performance indicators. |\",\"canGuestDownload\":true,\"createdDate\":\"2024-10-03T00:40:13.315Z\",\"description\":\"$b2\",\"explainability\":\"$b3\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$b4\",\"safetyAndSecurity\":\"| Field | Response |\\n|-------|----------|\\n| Model Application(s): | Speech Enhancement |\\n| Describe the life critical impact (if present). | Not Applicable for licensed use cases per [Maxine Evaluation EULA](https://developer.download.nvidia.com/maxine/nvidia-maxine-evaluation-license-24oct2023.pdf) |\\n| Use Case Restrictions: | Abide by [Maxine Evaluation EULA](https://developer.download.nvidia.com/maxine/nvidia-maxine-evaluation-license-24oct2023.pdf) |\\n| Model and dataset restrictions: | The Principle of Least Privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. |\",\"updatedDate\":\"2025-03-21T05:36:05.335Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"ce615fc2-1343-4b58-968e-17069d60b156\"}},\"spec\":{\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-21T05:36:05.946Z\",\"nvcfFunctionId\":\"3f0aeba3-6d91-4465-b8cc-cc2aef355186\",\"createdDate\":\"2024-10-03T00:40:13.518Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.nvidia.com/nim/maxine/studio-voice/latest/index.html\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: Your use of this API is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/maxine-studio-voice\"},\"usage\":\"$b5\",\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$b6\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$b7\"}]},\"artifactName\":\"studiovoice\"},\"config\":{\"name\":\"studiovoice\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"llama-3_2-nv-embedqa-1b-v2\",\"displayName\":\"llama-3.2-nv-embedqa-1b-v2\",\"publisher\":\"nvidia\",\"shortDescription\":\"Multilingual and cross-lingual text question-answering retrieval with long context support and optimized data storage efficiency.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llama-3_2-nv-embedqa-1b-v2.jpg\",\"labels\":[\"embedding\",\"nemo retriever\",\"run on rtx\",\"Retrieval Augmented Generation\",\"Text-to-Embedding\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"| Field | Response |\\n| ----- | ----- |\\n| Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing | None |\\n| Measures taken to mitigate against unwanted bias | None |\",\"canGuestDownload\":true,\"createdDate\":\"2024-12-16T21:01:03.160Z\",\"description\":\"$b8\",\"explainability\":\"$b9\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$ba\",\"safetyAndSecurity\":\"| Field | Response |\\n| ----- | ----- |\\n| Model Application(s): | Text Embedding for Retrieval |\\n| Describe the physical safety impact (if present). | Not Applicable |\\n| Use Case Restrictions: | Abide by [NVIDIA AI Foundation Models Community License Agreement](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). |\\n| Model and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. |\",\"updatedDate\":\"2025-03-24T23:53:37.185Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"80aa7b84-0bb6-43b3-b584-fe4a0bec7a1c\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nvidia/llama-3.2-nv-embedqa-1b-v2\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_2-nv-embedqa-1b-v2 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1\"}],\"paths\":{\"/embeddings\":{\"post\":{\"tags\":[\"Embeddings\"],\"summary\":\"Creates an embedding vector from the input text.\",\"operationId\":\"create_embedding\",\"requestBody\":{\"required\":true,\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"input\":{\"description\":\"Input text to embed. Max length is 8192 tokens.\",\"oneOf\":[{\"type\":\"string\"},{\"items\":{\"type\":\"string\"},\"type\":\"array\"}],\"minLength\":1,\"maxLength\":8192,\"title\":\"Input\"},\"model\":{\"type\":\"string\",\"description\":\"ID of the embedding model.\",\"example\":\"nvidia/llama-3.2-nv-embedqa-1b-v2\",\"default\":\"nvidia/llama-3.2-nv-embedqa-1b-v2\",\"title\":\"Model\"},\"input_type\":{\"type\":\"string\",\"enum\":[\"passage\",\"query\"],\"description\":\"nvidia/llama-3.2-nv-embedqa-1b-v2 operates in `passage` or `query` mode, and thus require the `input_type` parameter. `passage` is used when generating embeddings during indexing. `query` is used when generating embeddings during querying. It is very important to use the correct `input_type`. Failure to do so will result in large drops in retrieval accuracy.\",\"title\":\"Input Type\"},\"encoding_format\":{\"type\":\"string\",\"description\":\"The format to return the embeddings in.\",\"enum\":[\"float\",\"base64\"],\"default\":\"float\",\"title\":\"Encoding Format\"},\"truncate\":{\"type\":\"string\",\"description\":\"Specifies how inputs longer than the maximum token length of the model are handled. Passing `START` discards the start of the input. `END` discards the end of the input. In both cases, input is discarded until the remaining input is exactly the maximum input token length for the model. If `NONE` is selected, when the input exceeds the maximum input token length an error will be returned.\",\"enum\":[\"NONE\",\"START\",\"END\"],\"default\":\"NONE\",\"title\":\"Truncate\"},\"user\":{\"type\":\"string\",\"description\":\"Not implemented, but provided for API compliance. This field is ignored.\",\"title\":\"User\"}},\"required\":[\"input\",\"model\"]}}}},\"x-nvai-meta\":{\"name\":\"Create Text Embedding\",\"description\":\"Generates an embedding vector from the provided text\\nusing a specified model. The embedding can be returned\\nin either float array or base64-encoded format.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Embedding vector for text input\",\"requestJson\":\"{\\n \\\"input\\\": \\\"What is the capital of France?\\\",\\n \\\"model\\\": \\\"nvidia/llama-3.2-nv-embedqa-1b-v2\\\",\\n \\\"input_type\\\": \\\"query\\\",\\n \\\"encoding_format\\\": \\\"float\\\",\\n \\\"truncate\\\": \\\"NONE\\\"\\n}\\n\",\"responseJson\":\"$bb\"}],\"templates\":[{\"title\":\"Synchronous requests\",\"requestEjs\":{\"curl\":\"curl -X POST https://integrate.api.nvidia.com/v1/embeddings \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"input\\\": [\\\"\u003c%- request.input.replaceAll('\\\"', '\\\\\\\\\\\"').replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\\\"],\\n \\\"model\\\": \\\"nvidia/llama-3.2-nv-embedqa-1b-v2\\\",\\n \\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\",\\n \\\"encoding_format\\\": \\\"\u003c%- request.encoding_format %\u003e\\\",\\n \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"\\n }'\\n\",\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n api_key=\\\"$NVIDIA_API_KEY\\\",\\n base_url=\\\"https://integrate.api.nvidia.com/v1\\\"\\n)\\n\\nresponse = client.embeddings.create(\\n input=[\u003c%- JSON.stringify(request.input) %\u003e],\\n model=\\\"nvidia/llama-3.2-nv-embedqa-1b-v2\\\",\\n encoding_format=\\\"\u003c%- request.encoding_format %\u003e\\\",\\n extra_body={\\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\", \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"}\\n)\\n\\nprint(response.data[0].embedding)\\n\",\"langchain\":\"from langchain_nvidia_ai_endpoints import NVIDIAEmbeddings\\n\\nclient = NVIDIAEmbeddings(\\n model=\\\"nvidia/llama-3.2-nv-embedqa-1b-v2\\\", \\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n truncate=\\\"\u003c%- request.truncate %\u003e\\\", \\n )\\n\\nembedding = client.embed_query(\u003c%- JSON.stringify(request.input) %\u003e)\\nprint(embedding)\\n\"},\"response\":\"$bc\"}]},\"responses\":{\"200\":{\"description\":\"Successful response\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\",\"example\":\"list\"},\"data\":{\"type\":\"array\",\"items\":{\"$ref\":\"#/components/schemas/EmbeddingObject\"}},\"model\":{\"type\":\"string\",\"example\":\"nvidia/llama-3.2-nv-embedqa-1b-v2\"},\"usage\":{\"type\":\"object\",\"description\":\"Number of tokens\",\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"example\":0},\"total_tokens\":{\"type\":\"integer\",\"example\":0}}}}}}}},\"400\":{\"description\":\"Bad request\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\"},\"message\":{\"type\":\"string\"},\"detail\":{\"type\":\"object\",\"additionalProperties\":true},\"type\":{\"type\":\"string\"}}}}}}}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"EmbeddingObject\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\",\"example\":\"embedding\"},\"embedding\":{\"oneOf\":[{\"items\":{\"type\":\"number\"},\"type\":\"array\",\"description\":\"The embedding vector as a list of floats. The length of the vector depends on the model.\"},{\"type\":\"string\",\"description\":\"The embedding vector as a Base64 string. The length of the string depends on the model.\"}]},\"index\":{\"type\":\"integer\",\"description\":\"The index of the embedding in the list of embeddings.\"}},\"required\":[\"object\",\"embedding\",\"index\"]}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:53:37.986Z\",\"nvcfFunctionId\":\"08d83ec3-339c-4923-b64c-2da76e684a14\",\"createdDate\":\"2024-12-16T21:01:03.611Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-llama-3_2-nv-embedqa-1b-v2\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Community Model License\u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_2/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eLlama 3.2 Community License Agreement\u003c/a\u003e. Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/nvidia/containers/llama-3.2-nv-embedqa-1b-v2\"},\"playground\":{\"type\":\"custom\",\"input\":{\"items\":[{\"key\":\"input\",\"type\":\"text-area\"},{\"key\":\"input_type\",\"type\":\"select\"},{\"key\":\"encoding_format\",\"type\":\"select\"},{\"key\":\"truncate\",\"type\":\"select\"}]},\"parameters\":{\"omitProperties\":[\"input_type\",\"encoding_format\",\"truncate\"]},\"output\":{\"items\":[{\"key\":\"data[0].embedding\",\"type\":\"code\",\"language\":\"javascript\",\"formatAsJson\":true}]},\"requestBody\":\"{ \\\"input\\\": [\\\"\u003c%- request.input %\u003e\\\"], \\\"model\\\": \\\"nvidia/llama-3.2-nv-embedqa-1b-v2\\\", \\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\", \\\"encoding_format\\\": \\\"\u003c%- request.encoding_format %\u003e\\\", \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\" }\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$bd\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$be\"}]},\"artifactName\":\"llama-3_2-nv-embedqa-1b-v2\"},\"config\":{\"name\":\"llama-3_2-nv-embedqa-1b-v2\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"nvclip\",\"displayName\":\"nvclip\",\"publisher\":\"nvidia\",\"shortDescription\":\"NV-CLIP is a multimodal embeddings model for image and text.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/nvclip.jpg\",\"labels\":[\"Computer vision\",\"NVIDIA NIM\",\"Run on rtx\",\"multimodal embeddings\",\"text and image\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"| Field | Response |\\n| -- | -- |\\n|Participation considerations from adversely impacted groups [(protected classes)](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None of the Above |\\n| Measures taken to mitigate against unwanted bias: | Not Applicable |\",\"canGuestDownload\":true,\"createdDate\":\"2024-06-13T20:35:52.208Z\",\"description\":\"$bf\",\"explainability\":\"$c0\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$c1\",\"safetyAndSecurity\":\"| Field | Response |\\n| -- | -- |\\n| Model Application(s): | Image Search |\\n| Describe the life-critical application (if present). | None: Not within Operational Design Domain |\\n| Use Case Restrictions: | Abide by [https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\\\"](https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/) |\\n| Describe access restrictions (if any): | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. |\",\"updatedDate\":\"2025-03-24T23:57:08.104Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"55baee58-48ce-46ff-ba9f-99341657c6c8\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nvidia/nvclip\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"NVIDIA AI Foundation Models Community License\",\"url\":\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1\"}],\"tags\":[{\"name\":\"NVIDIA NV-CLIP API\",\"description\":\"Get a vector representation of a given input that can be easily consumed by machine learning models and algorithms.\"}],\"paths\":{\"/embeddings\":{\"post\":{\"tags\":[\"Embeddings\"],\"summary\":\"Creates an embedding vector representing the input text or image.\",\"description\":\"Invokes inference using the embedding parameters. If uploading large images, this POST should be used in conjunction with the NVCF API which allows for the upload of large assets. \\nYou can find details on how to use NVCF Asset APIs here: https://docs.api.nvidia.com/cloud-functions/reference/createasset.\",\"operationId\":\"createEmbedding\",\"parameters\":[{\"in\":\"header\",\"name\":\"NVCF-INPUT-ASSET-REFERENCES\",\"schema\":{\"type\":\"string\",\"maxLength\":370,\"format\":\"uuid\"},\"required\":false,\"description\":\"String of asset IDs separated by commas. Data is uploaded to AWS S3 using NVCF Asset APIs and associated with these asset IDs.If the size of an image is more than 180KB, it needs to be uploaded to a presigned S3 URL bucket. The presigned URL allows for secure and temporary access to the S3 bucket for uploading the image. Once the asset is requested, an asset ID is generated for it. Please include this asset ID in this header and to use the uploaded image in a prompt, you need to refer to it using the following format: `\u003cimg src=\\\"data:image/png;asset_id,{asset_id}\\\" /\u003e`.\"}],\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/EmbeddingsRequest\"}}},\"required\":true},\"x-nvai-meta\":{\"name\":\"Create Text Embedding\",\"description\":\"Generates an embedding vector from the provided text\\nusing a specified model. The embedding can be returned\\nin either float array or base64-encoded format.\\n\",\"path\":\"create\",\"templates\":[{\"title\":\"Synchronous requests\",\"requestEjs\":{\"curl\":\"$c2\",\"python\":\"$c3\",\"node\":\"$c4\"},\"response\":\"$c5\"}]},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/EmbeddingsResponse\"}}}},\"402\":{\"description\":\"Payment Required.\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"detail\":{\"type\":\"string\",\"maxLength\":256,\"format\":\"^[a-zA-Z-]{1,64}$\",\"description\":\"Contains specific information related to the error and why it occurred.\",\"example\":\"You have reached your limit of credits.\"}},\"type\":\"object\",\"title\":\"PaymentRequiredError\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/bc205f8e-1740-40df-8d32-c4321763498a\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/Errors\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/bc205f8e-1740-40df-8d32-c4321763498a\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"EmbeddingsRequest\":{\"additionalProperties\":false,\"properties\":{\"input\":{\"oneOf\":[{\"type\":\"string\",\"title\":\"string\",\"example\":\"This is a test.\",\"pattern\":\"^.*$\"},{\"type\":\"array\",\"items\":{\"type\":\"string\",\"pattern\":\"^.*$\"},\"maxItems\":64}],\"title\":\"Input\",\"description\":\"The list of images or texts that you want to generate embeddings for. Images should be in form of `data:image/{format};base64,{base64encodedimage}`. If the size of an image is more than 200KB, it needs to be uploaded to a presigned S3 bucket using NVCF Asset APIs. Once uploaded you can refer to it using the following format: `\u003cimg src=\\\"data:image/png;asset_id,{asset_id}\\\" /\u003e`. Accepted formats are `jpg`, `png` and `jpeg`.\",\"examples\":[[\"The quick brown fox jumped over the lazy dog\",\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAIAAACQd1PeAAAAEElEQVR4nGK6HcwNCAAA//8DTgE8HuxwEQAAAABJRU5ErkJggg==\"],[\"The quick brown fox jumped over the lazy dog\",\"data:image/png;asset_id,87b132b0-08f9-43ea-8fab-f107350a5d00\"]]},\"encoding_format\":{\"type\":\"string\",\"enum\":[\"float\",\"base64\"],\"title\":\"Encoding Format\",\"description\":\"The format to return the embeddings in. Can be either `float` or `base64`.\",\"default\":\"float\"},\"model\":{\"type\":\"string\",\"enum\":[\"nvidia/nvclip\"],\"title\":\"Model\",\"description\":\"ID of the embedding model.\",\"examples\":\"nvidia/nvclip\"},\"dimensions\":{\"description\":\"Not implemented, but provided for API compliance. This field is ignored.\\n\",\"type\":\"integer\",\"minimum\":1,\"maximum\":1,\"format\":\"int32\"},\"user\":{\"type\":\"string\",\"description\":\"Not implemented, but provided for API compliance. This field is ignored.\",\"pattern\":\"^.*$\",\"maxLength\":204800,\"example\":\"user123\"}},\"type\":\"object\",\"required\":[\"input\",\"model\"],\"title\":\"EmbeddingsRequest\"},\"EmbeddingsResponse\":{\"type\":\"object\",\"properties\":{\"object\":{\"description\":\"The object type, which is always `list`.\",\"enum\":[\"list\"],\"title\":\"Object\",\"type\":\"string\"},\"data\":{\"items\":{\"$ref\":\"#/components/schemas/Embedding\"},\"type\":\"array\",\"title\":\"Data\",\"description\":\"The list of embeddings generated by the model.\",\"maxItems\":128,\"examples\":[[{\"embedding\":[0.1,0.1,0.1],\"index\":0,\"object\":\"embedding\"},{\"embedding\":[0.1,0.1,0.1],\"index\":1,\"object\":\"embedding\"}]]},\"model\":{\"type\":\"string\",\"description\":\"Model used to generate embeddings.\",\"example\":\"nvidia/nvclip\",\"title\":\"Model\",\"const\":\"nvidia/nvclip\"},\"usage\":{\"allOf\":[{\"$ref\":\"#/components/schemas/Usage\"}],\"description\":\"Usage statistics for the embeddings request.\",\"examples\":[{\"num_images\":1,\"prompt_tokens\":80,\"total_tokens\":80}]}},\"required\":[\"object\",\"data\",\"usage\",\"model\"],\"title\":\"EmbeddingsResponse\"},\"Embedding\":{\"type\":\"object\",\"description\":\"Represents an embedding vector returned by embedding endpoint.\",\"properties\":{\"object\":{\"description\":\"The object type, which is always `embedding`.\",\"enum\":[\"embedding\"],\"title\":\"Object\",\"type\":\"string\"},\"index\":{\"type\":\"integer\",\"format\":\"int32\",\"title\":\"Index\",\"description\":\"The index of the embedding in the list of embeddings.\",\"minimum\":0,\"maximum\":127},\"embedding\":{\"oneOf\":[{\"type\":\"array\",\"description\":\"The embedding vector as a list of floats, The length of the vector depends on the model.\\n\",\"items\":[{\"type\":\"number\"}],\"maxItems\":1024},{\"type\":\"string\",\"description\":\"The embedding vector as base64 string. The length of the vector depends on the model.\\n\",\"format\":\"^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{4}|[A-Za-z0-9+/]{3}=|[A-Za-z0-9+/]{2}==)$\"}]},\"title\":\"Embedding\"},\"required\":[\"index\",\"object\",\"embedding\"],\"title\":\"Embeddings\"},\"Usage\":{\"properties\":{\"num_images\":{\"type\":\"integer\",\"format\":\"int32\",\"title\":\"Num Images\",\"minimum\":0,\"maximum\":64,\"description\":\"Number of images passed.\"},\"prompt_tokens\":{\"type\":\"integer\",\"format\":\"int32\",\"title\":\"Prompt Tokens\",\"description\":\"Number of tokens in the prompt.\",\"minimum\":77,\"maximum\":4928},\"total_tokens\":{\"type\":\"integer\",\"format\":\"int32\",\"title\":\"Total Tokens\",\"description\":\"Total number of tokens used in the request.\",\"minimum\":77,\"maximum\":4928}},\"type\":\"object\",\"required\":[\"num_images\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\"},\"Errors\":{\"properties\":{\"type\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"format\":\"^.{1, 128}$\",\"maxLength\":128,\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"format\":\"int32\",\"minimum\":100,\"maximum\":999,\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"format\":\"^.{1, 1024}$\",\"maxLength\":1024,\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"format\":\"^.{1, 256}$\",\"maxLength\":256,\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"maxLength\":36,\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-24T23:57:08.915Z\",\"nvcfFunctionId\":\"3072eebf-b0f0-4318-a5b8-5a45cd035b95\",\"createdDate\":\"2024-06-13T20:35:52.529Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-nvclip\",\"termsOfUse\":\"GOVERNING TERMS: GOVERNING TERMS: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://docs.nvidia.com/ai-foundation-models-community-license.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA AI Foundation Models Community License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"type\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"projects\":[{\"name\":\"Multimodal Search with NV-CLIP NIM\",\"url\":\"https://github.com/NVIDIA/metropolis-nim-workflows/tree/main/nim_workflows/nvclip_multimodal_search\",\"imageUrl\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/github-logo.jpg\",\"workbench\":false}],\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$c6\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$c7\"}]},\"artifactName\":\"nvclip\"},\"config\":{\"name\":\"nvclip\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"paddleocr\",\"displayName\":\"paddleocr\",\"publisher\":\"baidu\",\"shortDescription\":\"Model for table extraction that receives an image as input, runs OCR on the image, and returns the text within the image and its bounding boxes.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/paddleocr.jpg\",\"labels\":[\"Optical Character Detection\",\"Optical Character Recognition\",\"Table Extraction\",\"data ingestion\",\"extraction\",\"nemo retriever\",\"run on rtx\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-11-22T17:15:01.232Z\",\"description\":\"$c8\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T00:02:36.704Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"dce1327f-040d-4a56-a7e9-b6f67693a0dd\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"PaddleOCR\",\"description\":\"PaddleOCR\",\"version\":\"0.2.1-rc0\"},\"paths\":{\"/v1/infer\":{\"post\":{\"summary\":\"Post Infer\",\"operationId\":\"post_infer_v1_infer_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/PaddleOCRRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/PaddleOCRResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"name\":\"Inference model on image\",\"returns\":\"Returns a object list.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Image from example\",\"input\":{\"image\":\"https://assets.ngc.nvidia.com/products/api-catalog/paddle/paddleocr1.png\"},\"requestJson\":\"$c9\",\"responseJson\":\"$ca\"},{\"name\":\"Image from example\",\"input\":{\"image\":\"https://assets.ngc.nvidia.com/products/api-catalog/paddle/paddleocr2.png\"},\"requestJson\":\"$cb\",\"responseJson\":\"$cc\"}],\"templates\":[{\"title\":\"Default\",\"requestEjs\":{\"python\":\"import requests, base64\\n\\ninvoke_url = \\\"https://ai.api.nvidia.com/v1/cv/baidu/paddleocr\\\"\\n\\nwith open(\\\"paddleocr1.png\\\", \\\"rb\\\") as f:\\n image_b64 = base64.b64encode(f.read()).decode()\\n\\nassert len(image_b64) \u003c 180_000, \\\\\\n \\\"To upload larger images, use the assets API (see docs)\\\"\\n\\nheaders = {\\n \\\"Authorization\\\": \\\"Bearer $NVIDIA_API_KEY\\\",\\n \\\"Accept\\\": \\\"application/json\\\"\\n}\\n\\npayload = {\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": f\\\"data:image/png;base64,{image_b64}\\\"\\n }\\n ]\\n}\\n\\nresponse = requests.post(invoke_url, headers=headers, json=payload)\\n\\nprint(response.json())\\n\",\"node.js\":\"import axios from 'axios';\\nimport { readFile } from 'node:fs/promises';\\n\\nconst invokeUrl = \\\"https://ai.api.nvidia.com/v1/cv/baidu/paddleocr\\\";\\n\\nconst headers = {\\n \\\"Authorization\\\": \\\"Bearer $NVIDIA_API_KEY\\\",\\n \\\"Accept\\\": \\\"application/json\\\"\\n};\\n\\nreadFile(\\\"paddleocr1.png\\\")\\n .then(data =\u003e {\\n const imageB64 = Buffer.from(data).toString('base64');\\n if (imageB64.length \u003e 180_000) {\\n throw new Error(\\\"To upload larger images, use the assets API (see docs)\\\");\\n }\\n\\n const payload = {\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": `data:image/png;base64,${imageB64}`\\n }\\n ]\\n };\\n\\n return axios.post(invokeUrl, payload, { headers: headers, responseType: 'json' });\\n })\\n .then(response =\u003e {\\n console.log(JSON.stringify(response.data));\\n })\\n .catch(error =\u003e {\\n console.error(error);\\n });\\n\",\"curl\":\"image_b64=$( base64 -i paddleocr1.png )\\n\\naccept_header='Accept: application/json'\\n\\n# Construct the JSON payload\\necho '{\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": \\\"data:image/png;base64,'\\\"${image_b64}\\\"'\\\"\\n }\\n ]\\n}' \u003e payload.json\\n\\ncurl https://ai.api.nvidia.com/v1/cv/baidu/paddleocr \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"$accept_header\\\" \\\\\\n -d @payload.json\\n\"},\"response\":\"$cd\"}]}}},\"/v1/health/live\":{\"get\":{\"summary\":\"Health Live\",\"description\":\"Return service liveness status\",\"operationId\":\"health_live_v1_health_live_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthLiveResponse\"}}}}}}},\"/v1/health/ready\":{\"get\":{\"summary\":\"Health Ready\",\"description\":\"Return service readiness status\",\"operationId\":\"health_ready_v1_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthReadyResponse\"}}}}}}},\"/v1/metrics\":{\"get\":{\"summary\":\"Metrics\",\"description\":\"Handler for metrics endpoint.\",\"operationId\":\"metrics_v1_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"string\",\"title\":\"Response Metrics V1 Metrics Get\"}}}}}}},\"/v1/license\":{\"get\":{\"summary\":\"License\",\"description\":\"Handler for license endpoint.\",\"operationId\":\"license_v1_license_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/LicenseEndpointModel\"}}}}}}},\"/v1/metadata\":{\"get\":{\"summary\":\"Metadata\",\"description\":\"Handler for metadata endpoint.\",\"operationId\":\"metadata_v1_metadata_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/MetadataEndpointModel\"}}}}}}},\"/v1/manifest\":{\"get\":{\"summary\":\"Manifest\",\"description\":\"Handler for the manifest endpoint.\",\"operationId\":\"manifest_v1_manifest_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ManifestEndpointModel\"}}}}}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthLiveResponse\":{\"properties\":{\"live\":{\"type\":\"boolean\",\"title\":\"Live\"}},\"type\":\"object\",\"required\":[\"live\"],\"title\":\"HealthLiveResponse\"},\"HealthReadyResponse\":{\"properties\":{\"ready\":{\"type\":\"boolean\",\"title\":\"Ready\"}},\"type\":\"object\",\"required\":[\"ready\"],\"title\":\"HealthReadyResponse\"},\"Image\":{\"properties\":{\"type\":{\"type\":\"string\",\"enum\":[\"image_url\"],\"const\":\"image_url\",\"title\":\"Type\",\"default\":\"image_url\"},\"image_url\":{\"$ref\":\"#/components/schemas/ImageUrl\"}},\"type\":\"object\",\"required\":[\"image_url\"],\"title\":\"Image\"},\"ImageUrl\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"The URL of the image to be processed.\"}},\"type\":\"object\",\"required\":[\"url\"],\"title\":\"ImageUrl\"},\"LicenseEndpointModel\":{\"properties\":{\"name\":{\"type\":\"string\",\"title\":\"Name\",\"description\":\"The name of the license for the NIM container.\"},\"path\":{\"type\":\"string\",\"title\":\"Path\",\"description\":\"The filepath within the container containing the license content.\"},\"sha\":{\"type\":\"string\",\"title\":\"Sha\",\"description\":\"A SHA1 hash of the license contents.\"},\"size\":{\"type\":\"integer\",\"title\":\"Size\",\"description\":\"The number of characters in the license content.\"},\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"The url where this license is hosted externally.\"},\"type\":{\"type\":\"string\",\"enum\":[\"file\"],\"const\":\"file\",\"title\":\"Type\",\"description\":\"The format of the license content.\"},\"content\":{\"type\":\"string\",\"title\":\"Content\",\"description\":\"The license text.\"}},\"type\":\"object\",\"required\":[\"name\",\"path\",\"sha\",\"size\",\"url\",\"type\",\"content\"],\"title\":\"LicenseEndpointModel\",\"description\":\"A model representing the license response.\"},\"ManifestEndpointModel\":{\"properties\":{\"manifest_file\":{\"type\":\"string\",\"title\":\"Manifest File\",\"description\":\"The content of the manifest file describing the required model artifacts.\"}},\"type\":\"object\",\"required\":[\"manifest_file\"],\"title\":\"ManifestEndpointModel\",\"description\":\"A model representing the manifest response.\"},\"Message\":{\"properties\":{\"role\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Role\",\"default\":\"\"},\"content\":{\"items\":{\"$ref\":\"#/components/schemas/Image\"},\"type\":\"array\",\"minItems\":1,\"title\":\"Content\"}},\"type\":\"object\",\"required\":[\"content\"],\"title\":\"Message\"},\"MetadataEndpointModel\":{\"properties\":{\"assetInfo\":{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"title\":\"Assetinfo\",\"description\":\"A list of required container assets excluding model artifacts\"},\"licenseInfo\":{\"$ref\":\"#/components/schemas/LicenseEndpointModel\",\"description\":\"The license info.\"},\"modelInfo\":{\"items\":{\"$ref\":\"#/components/schemas/ModelInfo\"},\"type\":\"array\",\"title\":\"Modelinfo\",\"description\":\"A list of models being served by the NIM.\"},\"version\":{\"type\":\"string\",\"title\":\"Version\",\"description\":\"The version of the NIM service.\"}},\"type\":\"object\",\"required\":[\"assetInfo\",\"licenseInfo\",\"modelInfo\",\"version\"],\"title\":\"MetadataEndpointModel\",\"description\":\"A model representing the metadata response.\"},\"ModelInfo\":{\"properties\":{\"modelUrl\":{\"type\":\"string\",\"title\":\"Modelurl\"},\"shortName\":{\"type\":\"string\",\"title\":\"Shortname\"}},\"type\":\"object\",\"required\":[\"modelUrl\",\"shortName\"],\"title\":\"ModelInfo\",\"description\":\"A model representing the model response.\"},\"PaddleOCRRequest\":{\"properties\":{\"model\":{\"type\":\"string\",\"enum\":[\"baidu/paddleocr\"],\"const\":\"baidu/paddleocr\",\"title\":\"Model Name\",\"default\":\"baidu/paddleocr\"},\"messages\":{\"items\":{\"$ref\":\"#/components/schemas/Message\"},\"type\":\"array\",\"maxItems\":1,\"minItems\":1,\"title\":\"Messages\"}},\"type\":\"object\",\"required\":[\"messages\"],\"title\":\"PaddleOCRRequest\"},\"PaddleOCRResponse\":{\"properties\":{\"object\":{\"type\":\"string\",\"enum\":[\"list\"],\"const\":\"list\",\"title\":\"Object\",\"default\":\"list\"},\"data\":{\"items\":{\"$ref\":\"#/components/schemas/PaddleOCRResponseItem\"},\"type\":\"array\",\"title\":\"Data\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"usage\":{\"anyOf\":[{\"type\":\"object\"},{\"type\":\"null\"}],\"title\":\"Usage\"}},\"type\":\"object\",\"required\":[\"data\",\"model\"],\"title\":\"PaddleOCRResponse\"},\"PaddleOCRResponseItem\":{\"properties\":{\"index\":{\"type\":\"integer\",\"title\":\"Index\"},\"content\":{\"type\":\"string\",\"title\":\"Content\"},\"object\":{\"type\":\"string\",\"enum\":[\"string\"],\"const\":\"string\",\"title\":\"Object\",\"default\":\"string\"}},\"type\":\"object\",\"required\":[\"index\",\"content\"],\"title\":\"PaddleOCRResponseItem\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T00:02:37.700Z\",\"nvcfFunctionId\":\"43ace38a-357d-4440-9385-10a9d4c9a581\",\"createdDate\":\"2024-11-22T17:15:01.822Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.nvidia.com/nim/ingestion/table-extraction/latest/api-reference.html\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA-Evaluation-License-Agreement.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Evaluation License Agreement\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://github.com/PaddlePaddle/PaddleOCR/blob/main/LICENSE\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eApache License Version 2.0\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$ce\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$cf\"}]},\"artifactName\":\"paddleocr\"},\"config\":{\"name\":\"paddleocr\",\"type\":\"model\"}},{\"endpoint\":{\"artifact\":{\"name\":\"nv-yolox-page-elements-v1\",\"displayName\":\"nv-yolox-page-elements-v1\",\"publisher\":\"nvidia\",\"shortDescription\":\"Model for object detection, fine-tuned to detect charts, tables, and titles in documents.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/nv-yolox-page-elements-v1.jpg\",\"labels\":[\"Chart Detection\",\"Data ingestion\",\"Object Detection\",\"Table Detection\",\"extraction\",\"nemo retriever\",\"run on rtx\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-11-22T17:15:01.485Z\",\"description\":\"$d0\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T00:03:00.623Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"280e6b44-1440-4099-b338-2211f2261255\"}},\"spec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NeMo Retriever YOLOX Structured Images v1\",\"description\":\"You Only Look Once (YOLO) anchor-free object detection of charts and tables in structured images.\",\"version\":\"0.2.1-rc0\"},\"paths\":{\"/v1/infer\":{\"post\":{\"summary\":\"Post V1 Infer\",\"description\":\"Detect charts and tables in the provided images.\",\"operationId\":\"post_v1_infer_v1_infer_post\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/YoloxRequest\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/YoloxResponse\"}}}},\"422\":{\"description\":\"Validation Error\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HTTPValidationError\"}}}}},\"x-nvai-meta\":{\"name\":\"Inference model on image\",\"returns\":\"Returns a object list.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Image from example\",\"input\":{\"image\":\"https://assets.ngc.nvidia.com/products/api-catalog/yolox/yolox1.jpg\"},\"requestJson\":\"$d1\",\"responseJson\":\"$d2\"},{\"name\":\"Image from example\",\"input\":{\"image\":\"https://assets.ngc.nvidia.com/products/api-catalog/yolox/yolox2.jpg\"},\"requestJson\":\"$d3\",\"responseJson\":\"$d4\"}],\"templates\":[{\"title\":\"Default\",\"requestEjs\":{\"python\":\"import requests, base64\\n\\ninvoke_url = \\\"https://ai.api.nvidia.com/v1/cv/nvidia/nv-yolox-page-elements-v1\\\"\\n\\nwith open(\\\"yolox1.png\\\", \\\"rb\\\") as f:\\n image_b64 = base64.b64encode(f.read()).decode()\\n\\nassert len(image_b64) \u003c 180_000, \\\\\\n \\\"To upload larger images, use the assets API (see docs)\\\"\\n\\nheaders = {\\n \\\"Authorization\\\": \\\"Bearer $NVIDIA_API_KEY\\\",\\n \\\"Accept\\\": \\\"application/json\\\"\\n}\\n\\npayload = {\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": f\\\"data:image/png;base64,{image_b64}\\\"\\n }\\n ]\\n}\\n\\nresponse = requests.post(invoke_url, headers=headers, json=payload)\\n\\nprint(response.json())\\n\",\"node.js\":\"import axios from 'axios';\\nimport { readFile } from 'node:fs/promises';\\n\\nconst invokeUrl = \\\"https://ai.api.nvidia.com/v1/cv/nvidia/nv-yolox-page-elements-v1\\\";\\n\\nconst headers = {\\n \\\"Authorization\\\": \\\"Bearer $NVIDIA_API_KEY\\\",\\n \\\"Accept\\\": \\\"application/json\\\"\\n};\\n\\nreadFile(\\\"yolox1.png\\\")\\n .then(data =\u003e {\\n const imageB64 = Buffer.from(data).toString('base64');\\n if (imageB64.length \u003e 180_000) {\\n throw new Error(\\\"To upload larger images, use the assets API (see docs)\\\");\\n }\\n\\n const payload = {\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": `data:image/png;base64,${imageB64}`\\n }\\n ]\\n };\\n\\n return axios.post(invokeUrl, payload, { headers: headers, responseType: 'json' });\\n })\\n .then(response =\u003e {\\n console.log(JSON.stringify(response.data));\\n })\\n .catch(error =\u003e {\\n console.error(error);\\n });\\n\",\"curl\":\"image_b64=$( base64 -i yolox1.png )\\n\\naccept_header='Accept: application/json'\\n\\n# Construct the JSON payload\\necho '{\\n \\\"input\\\": [\\n {\\n \\\"type\\\": \\\"image_url\\\",\\n \\\"url\\\": \\\"data:image/png;base64,'\\\"${image_b64}\\\"'\\\"\\n }\\n ]\\n}' \u003e payload.json\\n\\ncurl https://ai.api.nvidia.com/v1/cv/nvidia/nv-yolox-page-elements-v1 \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"$accept_header\\\" \\\\\\n -d @payload.json\\n\"},\"response\":\"$d5\"}]}}},\"/v1/health/live\":{\"get\":{\"summary\":\"Health Live\",\"description\":\"Check if the service is running.\",\"operationId\":\"health_live_v1_health_live_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthLiveResponse\"}}}}}}},\"/v1/health/ready\":{\"get\":{\"summary\":\"Health Ready\",\"description\":\"Check if the service is ready to recieve traffic.\",\"operationId\":\"health_ready_v1_health_ready_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/HealthReadyResponse\"}}}}}}},\"/v1/metrics\":{\"get\":{\"summary\":\"Metrics\",\"description\":\"Handler for metrics endpoint.\",\"operationId\":\"metrics_v1_metrics_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"string\",\"title\":\"Response Metrics V1 Metrics Get\"}}}}}}},\"/v1/license\":{\"get\":{\"summary\":\"License\",\"description\":\"Handler for license endpoint.\",\"operationId\":\"license_v1_license_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/LicenseEndpointModel\"}}}}}}},\"/v1/metadata\":{\"get\":{\"summary\":\"Metadata\",\"description\":\"Handler for metadata endpoint.\",\"operationId\":\"metadata_v1_metadata_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/MetadataEndpointModel\"}}}}}}},\"/v1/manifest\":{\"get\":{\"summary\":\"Manifest\",\"description\":\"Handler for the manifest endpoint.\",\"operationId\":\"manifest_v1_manifest_get\",\"responses\":{\"200\":{\"description\":\"Successful Response\",\"content\":{\"application/json\":{\"schema\":{\"$ref\":\"#/components/schemas/ManifestEndpointModel\"}}}}}}}},\"components\":{\"schemas\":{\"HTTPValidationError\":{\"properties\":{\"detail\":{\"items\":{\"$ref\":\"#/components/schemas/ValidationError\"},\"type\":\"array\",\"title\":\"Detail\"}},\"type\":\"object\",\"title\":\"HTTPValidationError\"},\"HealthLiveResponse\":{\"properties\":{\"live\":{\"type\":\"boolean\",\"title\":\"Live\"}},\"type\":\"object\",\"required\":[\"live\"],\"title\":\"HealthLiveResponse\"},\"HealthReadyResponse\":{\"properties\":{\"ready\":{\"type\":\"boolean\",\"title\":\"Ready\"}},\"type\":\"object\",\"required\":[\"ready\"],\"title\":\"HealthReadyResponse\"},\"LicenseEndpointModel\":{\"properties\":{\"name\":{\"type\":\"string\",\"title\":\"Name\",\"description\":\"The name of the license for the NIM container.\"},\"path\":{\"type\":\"string\",\"title\":\"Path\",\"description\":\"The filepath within the container containing the license content.\"},\"sha\":{\"type\":\"string\",\"title\":\"Sha\",\"description\":\"A SHA1 hash of the license contents.\"},\"size\":{\"type\":\"integer\",\"title\":\"Size\",\"description\":\"The number of characters in the license content.\"},\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"The url where this license is hosted externally.\"},\"type\":{\"type\":\"string\",\"enum\":[\"file\"],\"const\":\"file\",\"title\":\"Type\",\"description\":\"The format of the license content.\"},\"content\":{\"type\":\"string\",\"title\":\"Content\",\"description\":\"The license text.\"}},\"type\":\"object\",\"required\":[\"name\",\"path\",\"sha\",\"size\",\"url\",\"type\",\"content\"],\"title\":\"LicenseEndpointModel\",\"description\":\"A model representing the license response.\"},\"ManifestEndpointModel\":{\"properties\":{\"manifest_file\":{\"type\":\"string\",\"title\":\"Manifest File\",\"description\":\"The content of the manifest file describing the required model artifacts.\"}},\"type\":\"object\",\"required\":[\"manifest_file\"],\"title\":\"ManifestEndpointModel\",\"description\":\"A model representing the manifest response.\"},\"MetadataEndpointModel\":{\"properties\":{\"assetInfo\":{\"items\":{\"type\":\"string\"},\"type\":\"array\",\"title\":\"Assetinfo\",\"description\":\"A list of required container assets excluding model artifacts\"},\"licenseInfo\":{\"$ref\":\"#/components/schemas/LicenseEndpointModel\",\"description\":\"The license info.\"},\"modelInfo\":{\"items\":{\"$ref\":\"#/components/schemas/ModelInfo\"},\"type\":\"array\",\"title\":\"Modelinfo\",\"description\":\"A list of models being served by the NIM.\"},\"version\":{\"type\":\"string\",\"title\":\"Version\",\"description\":\"The version of the NIM service.\"}},\"type\":\"object\",\"required\":[\"assetInfo\",\"licenseInfo\",\"modelInfo\",\"version\"],\"title\":\"MetadataEndpointModel\",\"description\":\"A model representing the metadata response.\"},\"ModelInfo\":{\"properties\":{\"modelUrl\":{\"type\":\"string\",\"title\":\"Modelurl\"},\"shortName\":{\"type\":\"string\",\"title\":\"Shortname\"}},\"type\":\"object\",\"required\":[\"modelUrl\",\"shortName\"],\"title\":\"ModelInfo\",\"description\":\"A model representing the model response.\"},\"Type\":{\"type\":\"string\",\"enum\":[\"chart\",\"table\",\"title\"],\"title\":\"Type\"},\"ValidationError\":{\"properties\":{\"loc\":{\"items\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"integer\"}]},\"type\":\"array\",\"title\":\"Location\"},\"msg\":{\"type\":\"string\",\"title\":\"Message\"},\"type\":{\"type\":\"string\",\"title\":\"Error Type\"}},\"type\":\"object\",\"required\":[\"loc\",\"msg\",\"type\"],\"title\":\"ValidationError\"},\"YoloxBoundingBox\":{\"properties\":{\"xmin\":{\"type\":\"number\",\"title\":\"Xmin\"},\"ymin\":{\"type\":\"number\",\"title\":\"Ymin\"},\"xmax\":{\"type\":\"number\",\"title\":\"Xmax\"},\"ymax\":{\"type\":\"number\",\"title\":\"Ymax\"},\"confidence\":{\"type\":\"number\",\"title\":\"Confidence\"}},\"type\":\"object\",\"required\":[\"xmin\",\"ymin\",\"xmax\",\"ymax\",\"confidence\"],\"title\":\"YoloxBoundingBox\"},\"YoloxImage\":{\"properties\":{\"type\":{\"type\":\"string\",\"enum\":[\"image_url\"],\"const\":\"image_url\",\"title\":\"Type\",\"default\":\"image_url\"},\"image_url\":{\"$ref\":\"#/components/schemas/YoloxImageUrl\"}},\"type\":\"object\",\"required\":[\"image_url\"],\"title\":\"YoloxImage\"},\"YoloxImageUrl\":{\"properties\":{\"url\":{\"type\":\"string\",\"title\":\"Url\",\"description\":\"The URL of the image to be processed.\"}},\"type\":\"object\",\"required\":[\"url\"],\"title\":\"YoloxImageUrl\"},\"YoloxMessage\":{\"properties\":{\"role\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Role\",\"default\":\"\"},\"content\":{\"items\":{\"$ref\":\"#/components/schemas/YoloxImage\"},\"type\":\"array\",\"minItems\":1,\"title\":\"Content\"}},\"type\":\"object\",\"required\":[\"content\"],\"title\":\"YoloxMessage\"},\"YoloxRequest\":{\"properties\":{\"model\":{\"type\":\"string\",\"enum\":[\"nvidia/nv-yolox-structured-images-v1\"],\"const\":\"nvidia/nv-yolox-structured-images-v1\",\"title\":\"Model\",\"default\":\"nvidia/nv-yolox-structured-images-v1\"},\"messages\":{\"items\":{\"$ref\":\"#/components/schemas/YoloxMessage\"},\"type\":\"array\",\"maxItems\":1,\"minItems\":1,\"title\":\"Messages\"},\"confidence_threshold\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}],\"title\":\"Confidence Threshold\",\"description\":\"Threshold for filtering low-confidence predictions\",\"default\":0.01},\"nms_threshold\":{\"anyOf\":[{\"type\":\"number\"},{\"type\":\"null\"}],\"title\":\"Nms Threshold\",\"description\":\"Threshold for filtering overlapping predictions via non-max suppression\",\"default\":0.5}},\"type\":\"object\",\"required\":[\"messages\"],\"title\":\"YoloxRequest\"},\"YoloxResponse\":{\"properties\":{\"object\":{\"type\":\"string\",\"enum\":[\"list\"],\"const\":\"list\",\"title\":\"Object\",\"default\":\"list\"},\"data\":{\"items\":{\"items\":{\"$ref\":\"#/components/schemas/YoloxResponseItem\"},\"type\":\"array\"},\"type\":\"array\",\"title\":\"Data\"},\"model\":{\"type\":\"string\",\"title\":\"Model\"},\"usage\":{\"anyOf\":[{\"type\":\"object\"},{\"type\":\"null\"}],\"title\":\"Usage\"}},\"type\":\"object\",\"required\":[\"data\",\"model\"],\"title\":\"YoloxResponse\"},\"YoloxResponseItem\":{\"properties\":{\"bboxes\":{\"items\":{\"$ref\":\"#/components/schemas/YoloxBoundingBox\"},\"type\":\"array\",\"title\":\"Bboxes\"},\"type\":{\"$ref\":\"#/components/schemas/Type\"}},\"type\":\"object\",\"required\":[\"bboxes\",\"type\"],\"title\":\"YoloxResponseItem\"}}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-25T00:03:03.075Z\",\"nvcfFunctionId\":\"96d723c1-5970-49a2-917a-e799e078242d\",\"createdDate\":\"2024-11-22T17:15:02.010Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.nvidia.com/nim/ingestion/object-detection/latest/api-reference.html\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA-Evaluation-License-Agreement.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA Evaluation License Agreement\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://github.com/Megvii-BaseDetection/YOLOX/blob/main/LICENSE\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eApache License Version 2.0\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$d6\"},{\"label\":\"Windows on RTX AI PCs (Beta)\",\"filename\":\"wsl2.md\",\"contents\":\"$d7\"}]},\"artifactName\":\"nv-yolox-page-elements-v1\"},\"config\":{\"name\":\"nv-yolox-page-elements-v1\",\"type\":\"model\"}}],\"items\":[\"$d8\",\"$d9\",\"$da\",\"$db\",\"$dc\",\"$dd\",\"$de\",\"$df\",\"$e0\"],\"params\":{},\"slotTitle\":[[\"$\",\"div\",null,{\"className\":\"mb-2 flex items-start gap-2 max-xs:justify-between\",\"children\":[[\"$\",\"h2\",null,{\"className\":\"text-ml font-medium leading-body tracking-less text-manitoulinLightWhite mb-0\",\"children\":\"AI Models for RTX AI PCs and Workstations\"}],[\"$\",\"$L26\",null,{\"href\":\"/models\",\"children\":[[\"$\",\"$L51\",null,{\"children\":[\"$\",\"svg\",\"arrow-right:fill\",{\"data-src\":\"https://brand-assets.cne.ngc.nvidia.com/assets/icons/3.1.0/fill/arrow-right.svg\",\"height\":\"1em\",\"width\":\"1em\",\"display\":\"inline-block\",\"data-icon-name\":\"arrow-right\",\"data-cache\":\"disabled\",\"color\":\"$undefined\",\"className\":\"btn-icon\"}]}],\"View All\"],\"className\":\"inline-flex items-center justify-center gap-2 text-center font-sans font-medium leading-text flex-row-reverse btn-tertiary btn-sm btn-pill text-nowrap mt-[3px]\"}]]}],[\"$\",\"p\",null,{\"className\":\"text-md font-normal text-manitoulinLightGray mb-0\",\"children\":\"Spanning language, speech, animation, content generation, and vision capabilities, run NVIDIA NIM microservices on your RTX AI PC.\"}],\" \"]}]\n"])</script><script>self.__next_f.push([1,"e1:I[3109,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"static/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"303\",\"static/chunks/app/(app-bar-layout)/explore/layout-bfa292509e5202a5.js\"],\"QuickActions\"]\n319:I[58596,[\"80650\",\"static/chunks/cef99e24-4550e53c8d136ca7.js\",\"71075\",\"static/chunks/a9670928-1c89a4311971ff31.js\",\"46781\",\"static/chunks/cb9c0786-198fb9a50c9ef41d.js\",\"30550\",\"static/chunks/e6046c14-e62a3eefe5b72ddd.js\",\"21610\",\"static/chunks/21610-7deb24ce331f2ac0.js\",\"13051\",\"static/chunks/13051-735cafa2c76b6612.js\",\"67406\",\"static/chunks/67406-3f94c755e5e6329f.js\",\"40459\",\"static/chunks/40459-2ff37b28b5fc1931.js\",\"10776\",\"static/chunks/10776-f8ba0fb8bdc5c1e0.js\",\"40726\",\"static/chunks/40726-b630e3e744bb7ac9.js\",\"60401\",\"stat"])</script><script>self.__next_f.push([1,"ic/chunks/60401-b0b40c42264c1e12.js\",\"70594\",\"static/chunks/70594-4a7d5e3efd2703ed.js\",\"58746\",\"static/chunks/58746-6d2311b776932c61.js\",\"16226\",\"static/chunks/16226-f37a631a5aab5d92.js\",\"39013\",\"static/chunks/39013-9472e74a6e6708c2.js\",\"83635\",\"static/chunks/83635-1a0e4c2cba2f2a76.js\",\"70953\",\"static/chunks/70953-edd738fa620d4ac1.js\",\"74846\",\"static/chunks/74846-2d3d81d1f42cfeb8.js\",\"7073\",\"static/chunks/7073-d4bd33567638e78d.js\",\"7465\",\"static/chunks/7465-cc6b0469e06e7696.js\",\"92340\",\"static/chunks/92340-f455198eac6d7201.js\",\"84784\",\"static/chunks/84784-71a92c03afc120a7.js\",\"82074\",\"static/chunks/82074-df560947c012b047.js\",\"25088\",\"static/chunks/25088-cd1635fc1fffeb0c.js\",\"4518\",\"static/chunks/4518-a3e00c8ba42e3b6c.js\",\"23362\",\"static/chunks/23362-ec1a0f0c6d917ab0.js\",\"90967\",\"static/chunks/90967-e90f8a4d90679f87.js\",\"4790\",\"static/chunks/4790-d19c0f2d8ca8f6d9.js\",\"303\",\"static/chunks/app/(app-bar-layout)/explore/layout-bfa292509e5202a5.js\"],\"SideBar\"]\ne5:[\"user\",\"assistant\"]\ne4:{\"description\":\"The role of the message author.\",\"enum\":\"$e5\",\"title\":\"Role\",\"type\":\"string\"}\ne8:{\"type\":\"string\"}\ne9:{\"type\":\"null\"}\ne7:[\"$e8\",\"$e9\"]\ne6:{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":\"$e7\"}\ne3:{\"role\":\"$e4\",\"content\":\"$e6\"}\nea:[\"role\",\"content\"]\ne2:{\"additionalProperties\":false,\"properties\":\"$e3\",\"required\":\"$ea\",\"title\":\"Message\",\"type\":\"object\"}\ned:{\"type\":\"string\",\"description\":\"Error type\"}\nee:{\"type\":\"string\",\"description\":\"Error title\"}\nef:{\"type\":\"integer\",\"description\":\"Error status code\"}\nf0:{\"type\":\"string\",\"description\":\"Detailed information about the error\"}\nf1:{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"}\nf2:{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}\nec:{\"type\":\"$ed\",\"title\":\"$ee\",\"status\":\"$ef\",\"detail\":\"$f0\",\"instance\":\"$f1\",\"requestId\":\"$f2\"}\nf3:[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"]\neb:{\"properties\":\"$ec\",\"type\":\"object\",\"required\":\"$f3\",\"title\":\"InvokeError\"}\nf4:T4bd,from openai import OpenAI\n\nclien"])</script><script>self.__next_f.push([1,"t = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\nf5:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request."])</script><script>self.__next_f.push([1,"messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();f6:T667,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e"])</script><script>self.__next_f.push([1,"\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003ef9:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\nfd:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\nff:[\"$e2\"]\n101:{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}\n100:[\"$101\"]\nfe:{\"allOf\":\"$ff\",\"description\":\"A chat completion message generated by the model.\",\"examples\":\"$100\"}\n105:[\"stop\",\"length\"]\n104:{\"enum\":\"$105\",\"type\":\"string\"}\n106:{\"type\":\"null\"}\n103:[\"$104\",\"$106\"]\n107:[\"stop\"]\n102:{\"anyOf\":\"$103\",\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":\"$107\",\"title\":\"Finish Reason\"}\nfc:{\"index\":\"$fd\",\"message\":\"$fe\",\"finish_reason\":\"$102\"}\n108:[\"index\",\"message\"]\nfb:{\"properties\":\"$fc\",\"required\":\"$108\",\"title\":\"Choice\",\"type\":\"object\"}\nfa:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$fb\",\"title\":\"Choices\",\"type\":\"array\"}\n10e:[25]\n10d:{\"description\":\"Number of tokens in the generated completion.\",\"examples\":\"$10e\",\"title\":\"Completion Tokens\",\"type\":\"integer\"}\n110:[9]\n10f:{\"description\":\"Number of tokens in the prompt.\",\"examples\":\"$110\",\"title\":\"Prompt Tokens\",\"type\":\"integer\"}\n112:[34]\n111:{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":\"$112\",\"title\":\"Total Tokens\",\"type\":\"integer\"}\n10c:{\"completion_tokens\":\"$10d\",\"prompt_tokens\":\"$10f\",\"total_tokens\":\"$111\"}\n113:[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"]\n10b:{\"properties\":\"$10c\",\"required\":\"$113\",\"title\":\"Usage\",\"type\":\"object\"}\n10a:[\"$10b\"]\n109:{\"allOf\":\"$10a\",\"description\":\"Usage statistics for the completion request.\"}\nf8:{\"id\":\"$f9\",\"choices\":\"$fa\",\"usage\":\"$109\"}\n114:[\"id\",\"choices\",\"usage\"]\nf7:{\"properties\":\"$"])</script><script>self.__next_f.push([1,"f8\",\"required\":\"$114\",\"title\":\"ChatCompletion\",\"type\":\"object\"}\n117:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\n11b:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\n11d:[\"$e2\"]\n11f:{\"content\":\"Ah,\",\"role\":\"assistant\"}\n11e:[\"$11f\"]\n11c:{\"allOf\":\"$11d\",\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":\"$11e\"}\n123:[\"stop\",\"length\"]\n122:{\"enum\":\"$123\",\"type\":\"string\"}\n124:{\"type\":\"null\"}\n121:[\"$122\",\"$124\"]\n120:{\"anyOf\":\"$121\",\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}\n11a:{\"index\":\"$11b\",\"delta\":\"$11c\",\"finish_reason\":\"$120\"}\n125:[\"index\",\"delta\"]\n119:{\"properties\":\"$11a\",\"required\":\"$125\",\"title\":\"ChoiceChunk\",\"type\":\"object\"}\n118:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$119\",\"title\":\"Choices\",\"type\":\"array\"}\n116:{\"id\":\"$117\",\"choices\":\"$118\"}\n126:[\"id\",\"choices\"]\n115:{\"properties\":\"$116\",\"required\":\"$126\",\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}\n129:{\"type\":\"string\",\"title\":\"Model\",\"default\":\"deepseek-ai/deepseek-r1\"}\n12d:{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}\n12c:[\"$12d\"]\n12b:[\"$12c\"]\n12a:{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":\"$12b\",\"items\":\"$e2\",\"title\":\"Messages\",\"type\":\"array\"}\n12e:{\"default\":0.6,\"description\":\"The sampling temperature to use for"])</script><script>self.__next_f.push([1," text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"}\n12f:{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"}\n130:{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"}\n131:{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"}\n132:{\"default\":4096,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"}\n133:{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"}\n137:{\"type\":\"string\"}\n136:{\"items\":\"$137\",\"type\":\"array\"}\n138:{\"type\":\"string\"}\n139:{\"type\":\"null\"}\n135:[\"$136\",\"$138\",\"$139\"]\n134:{\"anyOf\":\"$135\",\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop "])</script><script>self.__next_f.push([1,"sequence.\"}\n128:{\"model\":\"$129\",\"messages\":\"$12a\",\"temperature\":\"$12e\",\"top_p\":\"$12f\",\"frequency_penalty\":\"$130\",\"presence_penalty\":\"$131\",\"max_tokens\":\"$132\",\"stream\":\"$133\",\"stop\":\"$134\"}\n13a:[\"messages\"]\n127:{\"additionalProperties\":false,\"properties\":\"$128\",\"required\":\"$13a\",\"title\":\"ChatRequest\",\"type\":\"object\"}\n13b:T485,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/deepseek-ai/deepseek-r1:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"deepseek-ai/deepseek-r1\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Which number is larger, 9.11 or 9.8?\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).13c:T138d,"])</script><script>self.__next_f.push([1,"**Model Overview**\n\n## Description:\nDeepSeek-R1 is a first-generation reasoning model trained using large-scale reinforcement learning (RL) to solve complex reasoning tasks across domains such as math, code, and language. The model leverages RL to develop reasoning capabilities, which are further enhanced through supervised fine-tuning (SFT) to improve readability and coherence. DeepSeek-R1 achieves state-of-the-art results in various benchmarks and offers both its base models and distilled versions for community use.\n\nThis model is ready for both research and commercial use.\nFor more details, visit the [DeepSeek website](https://www.deepseek.com/).\n\n\u003cdiv align=\"center\"\u003e\n\u003cimg src=\"https://huggingface.co/deepseek-ai/DeepSeek-R1/resolve/main/figures/benchmark.jpg\" width=\"80%\" alt=\"Benchmarking\"/\u003e\n\u003c/div\u003e\n\n## Third-Party Community Consideration:\nThis model is not owned or developed by NVIDIA. It is a community-driven model created by DeepSeek AI. See the official [DeepSeek-R1 Model Card](https://huggingface.co/deepseek-ai/DeepSeek-R1) on Hugging Face for further details.\n\n## License/Terms of Use:\nGOVERNING TERMS: This trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Use of this model is governed by the [NVIDIA Community Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/). Additional Information: [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md).\n\n## References:\n- [DeepSeek GitHub Repository](https://github.com/deepseek-ai/DeepSeek-R1)\n- [DeepSeek-R1 Paper](https://arxiv.org/abs/2501.12948)\n\n## Model Architecture:\n\n**Architecture Type:** Mixture of Experts (MoE) \u003cbr\u003e\n**Network Architecture:** \u003cbr\u003e\n- Base Model: DeepSeek-V3-Base\n- Activated Parameters: 37 billion\n- Total Parameters: 671 billion\n- Distilled Models: Smaller, fine-tuned versions based on Qwen and Llama architectures.\n- Context Length: 128K tokens\n\n## Input:\n\n**Input Type(s):** Text \u003cbr\u003e\n**Input Format(s):** String \u003cbr\u003e\n**Input Parameters:** (1D) \u003cbr\u003e\n**Other Properties Related to Input:** \u003cbr\u003e\nDeepSeek recommends adhering to the following configurations when utilizing the DeepSeek-R1 series models, including benchmarking, to achieve the expected performance:\n\n1. Set the temperature within the range of 0.5-0.7 (0.6 is recommended) to prevent endless repetitions or incoherent outputs.\n2. **Avoid adding a system prompt; all instructions should be contained within the user prompt.**\n3. For mathematical problems, it is advisable to include a directive in your prompt such as: \"Please reason step by step, and put your final answer within \\boxed{}.\"\n4. When evaluating model performance, it is recommended to conduct multiple tests and average the results.\n\n## Output:\n**Output Type(s):** Text \u003cbr\u003e\n**Output Format:** String \u003cbr\u003e\n**Output Parameters:** (1D) \u003cbr\u003e\n\n## Software Integration:\n**Runtime Engine(s):** vLLM and SGLang \u003cbr\u003e\n**Supported Hardware Microarchitecture Compatibility:** NVIDIA Ampere, NVIDIA Blackwell, NVIDIA Jetson, NVIDIA Hopper, NVIDIA Lovelace, NVIDIA Pascal, NVIDIA Turing, and NVIDIA Volta architectures \u003cbr\u003e\n**[Preferred/Supported] Operating System(s):** Linux\n\n## Model Version(s):\nDeepSeek-R1 V1.0\n\n## Training, Testing, and Evaluation Datasets:\n### Training Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n### Testing Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n### Evaluation Dataset:\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n\n## Inference:\n**Engine:** SGLang\n**Test Hardware:** NVIDIA Hopper\n\n## Ethical Considerations:\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n## Model Limitations:\nThe base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive."])</script><script>self.__next_f.push([1,"13f:{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"}\n140:{\"name\":\"MIT License\",\"url\":\"https://huggingface.co/deepseek-ai/deepseek-r1/blob/main/LICENSE\"}\n13e:{\"title\":\"NVIDIA NIM API for deepseek-ai/deepseek-r1\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":\"$13f\",\"license\":\"$140\"}\n142:{\"url\":\"https://integrate.api.nvidia.com/v1/\"}\n141:[\"$142\"]\n146:[\"Chat\"]\n149:{\"schema\":\"$127\"}\n148:{\"application/json\":\"$149\"}\n147:{\"content\":\"$148\",\"required\":true}\n14d:{\"schema\":\"$f7\"}\n14e:{\"schema\":\"$115\"}\n14c:{\"application/json\":\"$14d\",\"text/event-stream\":\"$14e\"}\n14b:{\"description\":\"Invocation is fulfilled\",\"content\":\"$14c\"}\n152:{}\n153:{}\n151:{\"example\":\"$152\",\"schema\":\"$153\"}\n150:{\"application/json\":\"$151\"}\n156:{\"type\":\"string\",\"format\":\"uuid\"}\n155:{\"description\":\"requestId required for pooling\",\"schema\":\"$156\"}\n158:{\"type\":\"string\"}\n157:{\"description\":\"Invocation status\",\"schema\":\"$158\"}\n154:{\"NVCF-REQID\":\"$155\",\"NVCF-STATUS\":\"$157\"}\n14f:{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":\"$150\",\"headers\":\"$154\"}\n15c:{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n15b:{\"schema\":\"$eb\",\"example\":\"$15c\"}\n15a:{\"application/json\":\"$15b\"}\n159:{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":\"$15a\"}\n160:{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n15f:{\"schema\":\"$eb\",\"example\":\"$160\"}\n15"])</script><script>self.__next_f.push([1,"e:{\"application/json\":\"$15f\"}\n15d:{\"description\":\"The invocation ended with an error.\",\"content\":\"$15e\"}\n14a:{\"200\":\"$14b\",\"202\":\"$14f\",\"422\":\"$159\",\"500\":\"$15d\"}\n163:{\"name\":\"Which number is larger, 9.11 or 9.8?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Which number is larger, 9.11 or 9.8?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The number 9.11 is...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n164:{\"name\":\"How many 'r's are in 'strawberry'?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"How many 'r's are in 'strawberry'?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"In the word strawberry...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n162:[\"$163\",\"$164\"]\n168:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (r"])</script><script>self.__next_f.push([1,"equest.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\n169:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tok"])</script><script>self.__next_f.push([1,"ens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();16a:T667,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003e167:{\"python\":\"$168\",\"langChain\":\"from langchain_nvidia_ai_endpoints import Ch"])</script><script>self.__next_f.push([1,"atNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"$169\",\"curl\":\"$16a\"}\n166:{\"title\":\"No Streaming\",\"requestEjs\":\"$167\",\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n165:[\"$166\"]\n161:{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":\"$162\",\"templates\":\"$165\"}\n145:{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":\"$146\",\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":\"$147\",\"responses\":\"$14a\",\"x-nvai-meta\":\"$161\"}\n144:{\"post\":\"$145\"}\n143:{\"/chat/completions\":\"$144\"}\n16d:[]\n16c:{\"Token\":\"$16d\"}\n16b:[\"$16c\"]\n170:{\"type\":\"http\",\"scheme\":\"bearer\"}\n16f:{\"Token\":\"$170\"}\n171:{\"Errors\":\"$eb\",\"ChatCompletion\":\"$f7\",\"ChatCompletionChunk\":\"$115\",\"ChatRequest\":\"$127\",\"Choice\":\"$fb\",\"C"])</script><script>self.__next_f.push([1,"hoiceChunk\":\"$119\",\"Message\":\"$e2\",\"Usage\":\"$10b\"}\n16e:{\"securitySchemes\":\"$16f\",\"schemas\":\"$171\"}\n13d:{\"openapi\":\"3.1.0\",\"info\":\"$13e\",\"servers\":\"$141\",\"paths\":\"$143\",\"security\":\"$16b\",\"components\":\"$16e\"}\n175:[\"system\",\"context\",\"user\",\"assistant\",\"tool\"]\n174:{\"description\":\"The role of the message author.\",\"enum\":\"$175\",\"title\":\"Role\",\"type\":\"string\"}\n178:{\"type\":\"string\"}\n179:{\"type\":\"null\"}\n177:[\"$178\",\"$179\"]\n176:{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":\"$177\"}\n17c:{\"type\":\"string\"}\n17d:{\"type\":\"null\"}\n17b:[\"$17c\",\"$17d\"]\n17a:{\"anyOf\":\"$17b\",\"title\":\"Tool Call Id\",\"description\":\"The id of the tool call.\"}\n183:{\"type\":\"string\",\"title\":\"Id\"}\n185:[\"function\"]\n184:{\"type\":\"string\",\"enum\":\"$185\",\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"}\n18a:{\"type\":\"string\"}\n18b:{\"type\":\"null\"}\n189:[\"$18a\",\"$18b\"]\n188:{\"anyOf\":\"$189\",\"title\":\"Name\"}\n18e:{\"type\":\"string\"}\n18f:{\"type\":\"null\"}\n18d:[\"$18e\",\"$18f\"]\n18c:{\"anyOf\":\"$18d\",\"title\":\"Arguments\"}\n187:{\"name\":\"$188\",\"arguments\":\"$18c\"}\n190:[\"name\",\"arguments\"]\n186:{\"properties\":\"$187\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$190\",\"title\":\"FunctionCall\"}\n182:{\"id\":\"$183\",\"type\":\"$184\",\"function\":\"$186\"}\n191:[\"function\"]\n181:{\"properties\":\"$182\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$191\",\"title\":\"ToolCall\"}\n180:{\"items\":\"$181\",\"type\":\"array\"}\n192:{\"type\":\"null\"}\n17f:[\"$180\",\"$192\"]\n17e:{\"anyOf\":\"$17f\",\"title\":\"Tool Calls\",\"description\":\"The tool(s) called by the model.\"}\n173:{\"role\":\"$174\",\"content\":\"$176\",\"tool_call_id\":\"$17a\",\"tool_calls\":\"$17e\"}\n193:[\"role\",\"content\"]\n172:{\"additionalProperties\":false,\"properties\":\"$173\",\"required\":\"$193\",\"title\":\"Message\",\"type\":\"object\"}\n196:{\"type\":\"string\",\"description\":\"Error type\"}\n197:{\"type\":\"string\",\"description\":\"Error title\"}\n198:{\"type\":\"integer\",\"description\":\"Error status code\"}\n199:{\"type\":\"string\",\"description\":\"Detailed information about the error\"}\n19a:{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"}\n19b:{\"type\":\""])</script><script>self.__next_f.push([1,"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}\n195:{\"type\":\"$196\",\"title\":\"$197\",\"status\":\"$198\",\"detail\":\"$199\",\"instance\":\"$19a\",\"requestId\":\"$19b\"}\n19c:[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"]\n194:{\"properties\":\"$195\",\"type\":\"object\",\"required\":\"$19c\",\"title\":\"InvokeError\"}\n19d:T8e4,"])</script><script>self.__next_f.push([1,"{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Tell me about Dumbledore.\"\n }\n ],\n \"model\": \"meta/llama-3.3-70b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"describe_harry_potter_character\",\n \"description\": \"Returns information and images of Harry Potter characters.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"name\": {\n \"type\": \"string\",\n \"enum\": [\n \"Harry James Potter\",\n \"Hermione Jean Granger\",\n \"Ron Weasley\",\n \"Fred Weasley\",\n \"George Weasley\",\n \"Bill Weasley\",\n \"Percy Weasley\",\n \"Charlie Weasley\",\n \"Ginny Weasley\",\n \"Molly Weasley\",\n \"Arthur Weasley\",\n \"Neville Longbottom\",\n \"Luna Lovegood\",\n \"Draco Malfoy\",\n \"Albus Percival Wulfric Brian Dumbledore\",\n \"Minerva McGonagall\",\n \"Remus Lupin\",\n \"Rubeus Hagrid\",\n \"Sirius Black\",\n \"Severus Snape\",\n \"Bellatrix Lestrange\",\n \"Lord Voldemort\",\n \"Cedric Diggory\",\n \"Nymphadora Tonks\",\n \"James Potter\"\n ],\n \"description\": \"Name of the Harry Potter character\"\n }\n },\n \"required\": [\n \"name\"\n ]\n }\n }\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"19e:T536,{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"What is the weather in Santa Clara, CA?\"\n }\n ],\n \"model\": \"meta/llama-3.3-70b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_current_weather\",\n \"description\": \"A tool that gets the current weather at a location, if one is specified, and defaults to the user's location.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"location\": {\n \"type\": \"string\",\n \"description\": \"The location to find the weather of, or if not provided, it's the default location.\"\n },\n \"unit\": {\n \"type\": \"string\",\n \"enum\": [\n \"u\",\n \"m\"\n ],\n \"description\": \"Whether to use SI or USCS units (celsius or fahrenheit). Infer this from the user's location.\"\n }\n }\n }\n }\n }\n ]\n}\n19f:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create"])</script><script>self.__next_f.push([1,"(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\n1a0:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();1a1:T66f,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta"])</script><script>self.__next_f.push([1,"/llama-3.3-70b-instruct\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003e1a4:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\n1a8:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\n1aa:[\"$172\"]\n1ac:{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}\n1ab:[\"$1ac\"]\n1a9:{\"allOf\":\"$1aa\",\"description\":\"A chat completion message generated by the model.\",\"examples\":\"$1ab\"}\n1b0:[\"stop\",\"length\",\"tool_calls\"]\n1af:{\"enum\":\"$1b0\",\"type\":\"string\"}\n1b1:{\"type\":\"null\"}\n1ae:[\"$1af\",\"$1b1\"]\n1b2:[\"stop\"]\n1ad:{\"anyOf\":"])</script><script>self.__next_f.push([1,"\"$1ae\",\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":\"$1b2\",\"title\":\"Finish Reason\"}\n1a7:{\"index\":\"$1a8\",\"message\":\"$1a9\",\"finish_reason\":\"$1ad\"}\n1b3:[\"index\",\"message\"]\n1a6:{\"properties\":\"$1a7\",\"required\":\"$1b3\",\"title\":\"Choice\",\"type\":\"object\"}\n1a5:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$1a6\",\"title\":\"Choices\",\"type\":\"array\"}\n1b9:[25]\n1b8:{\"description\":\"Number of tokens in the generated completion.\",\"examples\":\"$1b9\",\"title\":\"Completion Tokens\",\"type\":\"integer\"}\n1bb:[9]\n1ba:{\"description\":\"Number of tokens in the prompt.\",\"examples\":\"$1bb\",\"title\":\"Prompt Tokens\",\"type\":\"integer\"}\n1bd:[34]\n1bc:{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":\"$1bd\",\"title\":\"Total Tokens\",\"type\":\"integer\"}\n1b7:{\"completion_tokens\":\"$1b8\",\"prompt_tokens\":\"$1ba\",\"total_tokens\":\"$1bc\"}\n1be:[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"]\n1b6:{\"properties\":\"$1b7\",\"required\":\"$1be\",\"title\":\"Usage\",\"type\":\"object\"}\n1b5:[\"$1b6\"]\n1b4:{\"allOf\":\"$1b5\",\"description\":\"Usage statistics for the completion request.\"}\n1a3:{\"id\":\"$1a4\",\"choices\":\"$1a5\",\"usage\":\"$1b4\"}\n1bf:[\"id\",\"choices\",\"usage\"]\n1a2:{\"properties\":\"$1a3\",\"required\":\"$1bf\",\"title\":\"ChatCompletion\",\"type\":\"object\"}\n1c2:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\n1c6:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\n1c8:[\"$172\"]\n1ca:{\"content\":\"Ah,\",\"role\":\"assistant\"}\n1c9:[\"$1ca\"]\n1c7:{\"allOf\":\"$1c8\",\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":\"$1c9\"}\n1ce:[\"stop\",\"length\",\"tool_calls\"]\n1cd:{\"enum\":\"$1ce\",\"type\":\"string\"}\n1cf:{\"type\":\"null\"}\n1cc:[\"$1cd\",\"$1cf\"]\n1cb:{\"anyOf\":\"$1cc\",\"default\":null,\"descri"])</script><script>self.__next_f.push([1,"ption\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}\n1c5:{\"index\":\"$1c6\",\"delta\":\"$1c7\",\"finish_reason\":\"$1cb\"}\n1d0:[\"index\",\"delta\"]\n1c4:{\"properties\":\"$1c5\",\"required\":\"$1d0\",\"title\":\"ChoiceChunk\",\"type\":\"object\"}\n1c3:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$1c4\",\"title\":\"Choices\",\"type\":\"array\"}\n1c1:{\"id\":\"$1c2\",\"choices\":\"$1c3\"}\n1d1:[\"id\",\"choices\"]\n1c0:{\"properties\":\"$1c1\",\"required\":\"$1d1\",\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}\n1d4:{\"type\":\"string\",\"title\":\"Model\",\"default\":\"meta/llama-3.3-70b-instruct\"}\n1d8:{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}\n1d7:[\"$1d8\"]\n1d6:[\"$1d7\"]\n1d5:{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":\"$1d6\",\"items\":\"$172\",\"title\":\"Messages\",\"type\":\"array\"}\n1d9:{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"}\n1da:{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"m"])</script><script>self.__next_f.push([1,"aximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"}\n1e1:[\"function\"]\n1e0:{\"type\":\"string\",\"enum\":\"$1e1\",\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"}\n1e4:{\"type\":\"string\",\"title\":\"Name\"}\n1e7:{\"type\":\"string\"}\n1e8:{\"type\":\"null\"}\n1e6:[\"$1e7\",\"$1e8\"]\n1e5:{\"anyOf\":\"$1e6\",\"title\":\"Description\"}\n1eb:{\"type\":\"object\"}\n1ec:{\"type\":\"null\"}\n1ea:[\"$1eb\",\"$1ec\"]\n1e9:{\"anyOf\":\"$1ea\",\"title\":\"Parameters\"}\n1e3:{\"name\":\"$1e4\",\"description\":\"$1e5\",\"parameters\":\"$1e9\"}\n1ed:[\"name\"]\n1e2:{\"properties\":\"$1e3\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$1ed\",\"title\":\"FunctionDefinition\"}\n1df:{\"type\":\"$1e0\",\"function\":\"$1e2\"}\n1ee:[\"function\"]\n1de:{\"properties\":\"$1df\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$1ee\",\"title\":\"ChatCompletionToolsParam\"}\n1dd:{\"items\":\"$1de\",\"type\":\"array\"}\n1ef:{\"type\":\"null\"}\n1dc:[\"$1dd\",\"$1ef\"]\n1db:{\"anyOf\":\"$1dc\",\"title\":\"Tools\"}\n1f0:{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"}\n1f1:{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"}\n1f2:{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"}\n1f3:{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"}\n1f7:{\"type\":\"string\"}\n1f6:{\"items\":\"$1f7\",\"type\":\"array\"}\n1f8:{\"type"])</script><script>self.__next_f.push([1,"\":\"string\"}\n1f9:{\"type\":\"null\"}\n1f5:[\"$1f6\",\"$1f8\",\"$1f9\"]\n1f4:{\"anyOf\":\"$1f5\",\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}\n1d3:{\"model\":\"$1d4\",\"messages\":\"$1d5\",\"temperature\":\"$1d9\",\"top_p\":\"$1da\",\"tools\":\"$1db\",\"frequency_penalty\":\"$1f0\",\"presence_penalty\":\"$1f1\",\"max_tokens\":\"$1f2\",\"stream\":\"$1f3\",\"stop\":\"$1f4\"}\n1fa:[\"messages\"]\n1d2:{\"additionalProperties\":false,\"properties\":\"$1d3\",\"required\":\"$1fa\",\"title\":\"ChatRequest\",\"type\":\"object\"}\n1ff:{\"type\":\"string\"}\n200:{\"type\":\"null\"}\n1fe:[\"$1ff\",\"$200\"]\n1fd:{\"anyOf\":\"$1fe\",\"title\":\"Name\"}\n1fc:{\"name\":\"$1fd\"}\n201:[\"name\"]\n1fb:{\"properties\":\"$1fc\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$201\",\"title\":\"ChatCompletionNamedFunction\"}\n202:T49d,## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Pull and Run the NIM\n\n```bash\n$ docker login nvcr.io\nUsername: $oauthtoken\nPassword: \u003cPASTE_API_KEY_HERE\u003e\n```\n\nPull and run the NVIDIA NIM with the command below. This will download the optimized model for your infrastructure.\n\n```bash\nexport NGC_API_KEY=\u003cPASTE_API_KEY_HERE\u003e\nexport LOCAL_NIM_CACHE=~/.cache/nim\nmkdir -p \"$LOCAL_NIM_CACHE\"\ndocker run -it --rm \\\n --gpus all \\\n --shm-size=16GB \\\n -e NGC_API_KEY \\\n -v \"$LOCAL_NIM_CACHE:/opt/nim/.cache\" \\\n -u $(id -u) \\\n -p 8000:8000 \\\n nvcr.io/nim/meta/llama-3.3-70b-instruct:latest\n```\n\n## Step 3. Test the NIM\n\nYou can now make a local API call using this curl command:\n\n```bash\ncurl -X 'POST' \\\n'http://0.0.0.0:8000/v1/chat/completions' \\\n-H 'accept: application/json' \\\n-H 'Content-Type: application/json' \\\n-d '{\n \"model\": \"meta/llama-3.3-70b-instruct\",\n \"messages\": [{\"role\":\"user\", \"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n \"max_tokens\": 64\n}'\n```\n\nFor more details on getting started with this NIM, visit the [NVIDIA NIM Docs](https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html).203:T517,### What is NVIDIA AI Wo"])</script><script>self.__next_f.push([1,"rkbench?\n\nNVIDIA AI Workbench is a free, lightweight developer platform for building and running AI projects seamlessly across GPU systems. [Learn More](https://www.nvidia.com/en-us/deep-learning-ai/solutions/data-science/workbench/)\n\n## Step 1. Generate API Key\n\n::generate-api-key\n\n## Step 2. Clone Example Project with AI Workbench\n\nClick [HERE](https://build.nvidia.com/open-ai-workbench/aHR0cHM6Ly9naXRodWIuY29tL05WSURJQS93b3JrYmVuY2gtZXhhbXBsZS1kb3dubG9hZGFibGUtbmlt) to clone the project to your AI Workbench on your system of choice.\n\n## Step 3. Run NIM with AI Workbench\n\n* Under **Environment \u003e Compose**, select the ``meta/llama-3.3-70b-instruct`` compose profile from the dropdown. \n\n* Select **Start** and wait for the container to be ready. \n\n* Click **Open Chat** on the top right. Enter your ``NVIDIA_API_KEY`` from Step 1 when prompted. \n\n* Select **Use a NIM on the Host GPU** in the browser chat app and enter your GPU details. Select ``meta/llama-3.3-70b-instruct`` and enter a query. \n\n### Chat with Build Endpoints\n\n* Click **Open Chat** on the top right. Enter your ``NVIDIA_API_KEY`` from Step 1 if prompted. \n\n* To use the Build endpoint, select **Use Remote Endpoints** in the browser chat app. Select ``meta/llama-3.3-70b-instruct`` and enter a query. 204:T4369,"])</script><script>self.__next_f.push([1,"## Model Information\n\nThe Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction-tuned text-only model is optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks.\n\nModels are accelerated by [TensorRT-LLM](https://github.com/NVIDIA/TensorRT-LLM), a library for optimizing Large Language Model (LLM) inference on NVIDIA GPUs.\n\nThis model is ready for commercial use.\n\n**Model Developer**: Meta\n\n**Model Architecture:** Llama 3.3 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety. \n\n| | Training Data | Params | Input modalities | Output modalities | Context length | GQA | Token count | Knowledge cutoff |\n| :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- | :---- |\n| Llama 3.3 (text only) | A new mix of publicly available online data. | 70B | Multilingual Text | Multilingual Text and code | 128k | Yes | 15T+ | December 2023 |\n\n**Supported languages:** English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.\n\n**Llama 3.3 model**. Token counts refer to pretraining data only. All model versions use Grouped-Query Attention (GQA) for improved inference scalability.\n\n**Model Release Date:** \n\n* **70B Instruct: December 6, 2024** \n\n**Status:** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback.\n\n**License** A custom commercial license, the Llama 3.3 Community License Agreement, is available at: [https://github.com/meta-llama/llama-models/blob/main/models/llama3\\_3/LICENSE](https://github.com/meta-llama/llama-models/blob/main/models/llama3_3/LICENSE)\n\n## Intended Use\n\n**Intended Use Cases** Llama 3.3 is intended for commercial and research use in multiple languages. Instruction tuned text only models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks. The Llama 3.3 model also supports the ability to leverage the outputs of its models to improve other models including synthetic data generation and distillation. The Llama 3.3 Community License allows for these use cases. \n\n**Out-of-scope** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in any other way that is prohibited by the Acceptable Use Policy and Llama 3.3 Community License. Use in languages beyond those explicitly referenced as supported in this model card\\*\\*.\n\n\\*\\*Note: Llama 3.3 has been trained on a broader collection of languages than the 8 supported languages. Developers may fine-tune Llama 3.3 models for languages beyond the 8 supported languages provided they comply with the Llama 3.3 Community License and the Acceptable Use Policy and in such cases are responsible for ensuring that any uses of Llama 3.3 in additional languages is done in a safe and responsible manner.\n\n## Hardware and Software\n\n**Training Factors** We used custom training libraries, Meta's custom built GPU cluster, and production infrastructure for pretraining. Fine-tuning, annotation, and evaluation were also performed on production infrastructure.\n\n**Training Energy Use** Training utilized a cumulative of **39.3**M GPU hours of computation on H100-80GB (TDP of 700W) type hardware, per the table below. Training time is the total GPU time required for training each model and power consumption is the peak power capacity per GPU device used, adjusted for power usage efficiency. \n\n## \n\n**Training Greenhouse Gas Emissions** Estimated total location-based greenhouse gas emissions were **11,390** tons CO2eq for training. Since 2020, Meta has maintained net zero greenhouse gas emissions in its global operations and matched 100% of its electricity use with renewable energy, therefore the total market-based greenhouse gas emissions for training were 0 tons CO2eq.\n\n| | Training Time (GPU hours) | Training Power Consumption (W) | Training Location-Based Greenhouse Gas Emissions (tons CO2eq) | Training Market-Based Greenhouse Gas Emissions (tons CO2eq) |\n| :---- | :---: | :---: | :---: | :---: |\n| Llama 3.3 70B | 7.0M | 700 | 2,040 | 0 |\n\nThe methodology used to determine training energy use and greenhouse gas emissions can be found [here](https://arxiv.org/pdf/2204.05149). Since Meta is openly releasing these models, the training energy use and greenhouse gas emissions will not be incurred by others.\n\n## Training Data\n\n**Overview:** Llama 3.3 was pretrained on \\~15 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over 25M synthetically generated examples. \n\n**Data Freshness:** The pretraining data has a cutoff of December 2023\\.\n\n## Benchmarks \\- English Text\n\nIn this section, we report the results for Llama 3.3 relative to our previous models. \n\n### Instruction Tuned models\n\n## \n\n| Category | Benchmark | \\# Shots | Metric | Llama 3.1 8B Instruct | Llama 3.1 70B Instruct | Llama-3.3 70B Instruct | Llama 3.1 405B Instruct |\n| :---- | :---- | ----- | :---- | ----- | ----- | ----- | ----- |\n| | MMLU (CoT) | 0 | macro\\_avg/acc | 73.0 | 86.0 | 86.0 | 88.6 |\n| | MMLU Pro (CoT) | 5 | macro\\_avg/acc | 48.3 | 66.4 | 68.9 | 73.3 |\n| Steerability | IFEval | | | 80.4 | 87.5 | 92.1 | 88.6 |\n| Reasoning | GPQA Diamond (CoT) | 0 | acc | 31.8 | 48.0 | 50.5 | 49.0 |\n| Code | HumanEval | 0 | pass@1 | 72.6 | 80.5 | 88.4 | 89.0 |\n| | MBPP EvalPlus (base) | 0 | pass@1 | 72.8 | 86.0 | 87.6 | 88.6 |\n| Math | MATH (CoT) | 0 | sympy\\_intersection\\_score | 51.9 | 68.0 | 77.0 | 73.8 |\n| Tool Use | BFCL v2 | 0 | overall\\_ast\\_summary/macro\\_avg/valid | 65.4 | 77.5 | 77.3 | 81.1 |\n| Multilingual | MGSM | 0 | em | 68.9 | 86.9 | 91.1 | 91.6 |\n\n## \n\n## Responsibility \u0026 Safety\n\nAs part of our Responsible release approach, we followed a three-pronged strategy to managing trust \u0026 safety risks:\n\n* Enable developers to deploy helpful, safe and flexible experiences for their target audience and for the use cases supported by Llama. \n* Protect developers against adversarial users aiming to exploit Llama capabilities to potentially cause harm. \n* Provide protections for the community to help prevent the misuse of our models.\n\n### Responsible deployment \n\nLlama is a foundational technology designed to be used in a variety of use cases, examples on how Meta’s Llama models have been responsibly deployed can be found in our [Community Stories webpage](https://llama.meta.com/community-stories/). Our approach is to build the most helpful models enabling the world to benefit from the technology power, by aligning our model safety for the generic use cases addressing a standard set of harms. Developers are then in the driver seat to tailor safety for their use case, defining their own policy and deploying the models with the necessary safeguards in their Llama systems. Llama 3.3 was developed following the best practices outlined in our Responsible Use Guide, you can refer to the [Responsible Use Guide](https://llama.meta.com/responsible-use-guide/) to learn more. \n\n#### Llama 3.3 Instruct \n\nOur main objectives for conducting safety fine-tuning are to provide the research community with a valuable resource for studying the robustness of safety fine-tuning, as well as to offer developers a readily available, safe, and powerful model for various applications to reduce the developer workload to deploy safe AI systems. For more details on the safety mitigations implemented please read the Llama 3 paper. \n\n**Fine-tuning data** \nWe employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks. We’ve developed many large language model (LLM)-based classifiers that enable us to thoughtfully select high-quality prompts and responses, enhancing data quality control. \n\n**Refusals and Tone** \nBuilding on the work we started with Llama 3, we put a great emphasis on model refusals to benign prompts as well as refusal tone. We included both borderline and adversarial prompts in our safety data strategy, and modified our safety data responses to follow tone guidelines. \n\n#### Llama 3.3 Systems\n\n**Large language models, including Llama 3.3, are not designed to be deployed in isolation but instead should be deployed as part of an overall AI system with additional safety guardrails as required.** Developers are expected to deploy system safeguards when building agentic systems. Safeguards are key to achieve the right helpfulness-safety alignment as well as mitigating safety and security risks inherent to the system and any integration of the model or system with external tools. \n\nAs part of our responsible release approach, we provide the community with [safeguards](https://llama.meta.com/trust-and-safety/) that developers should deploy with Llama models or other LLMs, including Llama Guard 3, Prompt Guard and Code Shield. All our [reference implementations](https://github.com/meta-llama/llama-agentic-system) demos contain these safeguards by default so developers can benefit from system-level safety out-of-the-box. \n\n#### Capability specific considerations \n\n**Tool-use**: Just like in standard software development, developers are responsible for the integration of the LLM with the tools and services of their choice. They should define a clear policy for their use case and assess the integrity of the third party services they use to be aware of the safety and security limitations when using this capability. Refer to the Responsible Use Guide for best practices on the safe deployment of the third party safeguards. \n\n**Multilinguality**: Llama 3.3 supports 7 languages in addition to English: French, German, Hindi, Italian, Portuguese, Spanish, and Thai. Llama may be able to output text in other languages than those that meet performance thresholds for safety and helpfulness. We strongly discourage developers from using this model to converse in non-supported languages without implementing finetuning and system controls in alignment with their policies and the best practices shared in the Responsible Use Guide. \n\n### Evaluations\n\nWe evaluated Llama models for common use cases as well as specific capabilities. Common use cases evaluations measure safety risks of systems for most commonly built applications including chat bot, coding assistant, tool calls. We built dedicated, adversarial evaluation datasets and evaluated systems composed of Llama models and Llama Guard 3 to filter input prompt and output response. It is important to evaluate applications in context, and we recommend building dedicated evaluation dataset for your use case. Prompt Guard and Code Shield are also available if relevant to the application. \n\nCapability evaluations measure vulnerabilities of Llama models inherent to specific capabilities, for which were crafted dedicated benchmarks including long context, multilingual, tools calls, coding or memorization.\n\n**Red Teaming** \nFor both scenarios, we conducted recurring red teaming exercises with the goal of discovering risks via adversarial prompting and we used the learnings to improve our benchmarks and safety tuning datasets. \n\nWe partnered early with subject-matter experts in critical risk areas to understand the nature of these real-world harms and how such models may lead to unintended harm for society. Based on these conversations, we derived a set of adversarial goals for the red team to attempt to achieve, such as extracting harmful information or reprogramming the model to act in a potentially harmful capacity. The red team consisted of experts in cybersecurity, adversarial machine learning, responsible AI, and integrity in addition to multilingual content specialists with background in integrity issues in specific geographic markets. . \n\n### Critical and Other Risks \n\n### We specifically focused our efforts on mitigating the following critical risk areas:\n\n**1\\. CBRNE (Chemical, Biological, Radiological, Nuclear, and Explosive materials) helpfulness** \nTo assess risks related to proliferation of chemical and biological weapons, we performed uplift testing designed to assess whether use of the Llama 3.3 model could meaningfully increase the capabilities of malicious actors to plan or carry out attacks using these types of weapons. \n\n**2\\. Child Safety**\n\nChild Safety risk assessments were conducted using a team of experts, to assess the model’s capability to produce outputs that could result in Child Safety risks and inform on any necessary and appropriate risk mitigations via fine tuning. We leveraged those expert red teaming sessions to expand the coverage of our evaluation benchmarks through Llama 3 model development. For Llama 3, we conducted new in-depth sessions using objective based methodologies to assess the model risks along multiple attack vectors including the additional languages Llama 3 is trained on. We also partnered with content specialists to perform red teaming exercises assessing potentially violating content while taking account of market specific nuances or experiences. \n\n**3\\. Cyber Attack Enablement** \nOur cyber attack uplift study investigated whether LLMs can enhance human capabilities in hacking tasks, both in terms of skill level and speed. \n\nOur attack automation study focused on evaluating the capabilities of LLMs when used as autonomous agents in cyber offensive operations, specifically in the context of ransomware attacks. This evaluation was distinct from previous studies that considered LLMs as interactive assistants. The primary objective was to assess whether these models could effectively function as independent agents in executing complex cyber-attacks without human intervention.\n\n### Community \n\nGenerative AI safety requires expertise and tooling, and we believe in the strength of the open community to accelerate its progress. We are active members of open consortiums, including the AI Alliance, Partnership on AI and MLCommons, actively contributing to safety standardization and transparency. We encourage the community to adopt taxonomies like the MLCommons Proof of Concept evaluation to facilitate collaboration and transparency on safety and content evaluations. Our Purple Llama tools are open sourced for the community to use and widely distributed across ecosystem partners including cloud service providers. We encourage community contributions to our [Github repository](https://github.com/meta-llama/PurpleLlama). \n\nWe also set up the [Llama Impact Grants](https://llama.meta.com/llama-impact-grants/) program to identify and support the most compelling applications of Meta’s Llama model for societal benefit across three categories: education, climate and open innovation. The 20 finalists from the hundreds of applications can be found [here](https://llama.meta.com/llama-impact-grants/#finalists). \n\nFinally, we put in place a set of resources including an [output reporting mechanism](https://developers.facebook.com/llama_output_feedback) and [bug bounty program](https://www.facebook.com/whitehat) to continuously improve the Llama technology with the help of the community.\n\n## Ethical Considerations and Limitations\n\nThe core values of Llama 3.3 are openness, inclusivity and helpfulness. It is meant to serve everyone, and to work for a wide range of use cases. It is thus designed to be accessible to people across many different backgrounds, experiences and perspectives. Llama 3.3 addresses users and their needs as they are, without insertion unnecessary judgment or normativity, while reflecting the understanding that even content that may appear problematic in some cases can serve valuable purposes in others. It respects the dignity and autonomy of all users, especially in terms of the values of free thought and expression that power innovation and progress. \n\nBut Llama 3.3 is a new technology, and like any new technology, there are risks associated with its use. Testing conducted to date has not covered, nor could it cover, all scenarios. For these reasons, as with all LLMs, Llama 3.3’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 3.3 model, developers should perform safety testing and tuning tailored to their specific applications of the model. Please refer to available resources including our [Responsible Use Guide](https://llama.meta.com/responsible-use-guide), [Trust and Safety](https://llama.meta.com/trust-and-safety/) solutions, and other [resources](https://llama.meta.com/docs/get-started/) to learn more about responsible development. \n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"207:{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"}\n208:{\"name\":\"Llama 3.3 License\",\"url\":\"https://www.llama.com/llama3_3/license/\"}\n206:{\"title\":\"NVIDIA NIM API for meta/llama-3.3-70b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/meta-llama-3_3-70b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":\"$207\",\"license\":\"$208\"}\n20a:{\"url\":\"https://integrate.api.nvidia.com/v1/\"}\n209:[\"$20a\"]\n20e:[\"Chat\"]\n211:{\"schema\":\"$1d2\"}\n210:{\"application/json\":\"$211\"}\n20f:{\"content\":\"$210\",\"required\":true}\n215:{\"schema\":\"$1a2\"}\n216:{\"schema\":\"$1c0\"}\n214:{\"application/json\":\"$215\",\"text/event-stream\":\"$216\"}\n213:{\"description\":\"Invocation is fulfilled\",\"content\":\"$214\"}\n21a:{}\n21b:{}\n219:{\"example\":\"$21a\",\"schema\":\"$21b\"}\n218:{\"application/json\":\"$219\"}\n21e:{\"type\":\"string\",\"format\":\"uuid\"}\n21d:{\"description\":\"requestId required for pooling\",\"schema\":\"$21e\"}\n220:{\"type\":\"string\"}\n21f:{\"description\":\"Invocation status\",\"schema\":\"$220\"}\n21c:{\"NVCF-REQID\":\"$21d\",\"NVCF-STATUS\":\"$21f\"}\n217:{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":\"$218\",\"headers\":\"$21c\"}\n224:{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n223:{\"schema\":\"$194\",\"example\":\"$224\"}\n222:{\"application/json\":\"$223\"}\n221:{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":\"$222\"}\n228:{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n227:{\"schema\":\"$194\",\"example\":\"$228\"}\n226:{\"appl"])</script><script>self.__next_f.push([1,"ication/json\":\"$227\"}\n225:{\"description\":\"The invocation ended with an error.\",\"content\":\"$226\"}\n212:{\"200\":\"$213\",\"202\":\"$217\",\"422\":\"$221\",\"500\":\"$225\"}\n22b:{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n22d:T8e4,"])</script><script>self.__next_f.push([1,"{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"Tell me about Dumbledore.\"\n }\n ],\n \"model\": \"meta/llama-3.3-70b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"describe_harry_potter_character\",\n \"description\": \"Returns information and images of Harry Potter characters.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"name\": {\n \"type\": \"string\",\n \"enum\": [\n \"Harry James Potter\",\n \"Hermione Jean Granger\",\n \"Ron Weasley\",\n \"Fred Weasley\",\n \"George Weasley\",\n \"Bill Weasley\",\n \"Percy Weasley\",\n \"Charlie Weasley\",\n \"Ginny Weasley\",\n \"Molly Weasley\",\n \"Arthur Weasley\",\n \"Neville Longbottom\",\n \"Luna Lovegood\",\n \"Draco Malfoy\",\n \"Albus Percival Wulfric Brian Dumbledore\",\n \"Minerva McGonagall\",\n \"Remus Lupin\",\n \"Rubeus Hagrid\",\n \"Sirius Black\",\n \"Severus Snape\",\n \"Bellatrix Lestrange\",\n \"Lord Voldemort\",\n \"Cedric Diggory\",\n \"Nymphadora Tonks\",\n \"James Potter\"\n ],\n \"description\": \"Name of the Harry Potter character\"\n }\n },\n \"required\": [\n \"name\"\n ]\n }\n }\n }\n ]\n}\n"])</script><script>self.__next_f.push([1,"22c:{\"name\":\"Tell me about Dumbledore.\",\"requestJson\":\"$22d\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n22f:T536,{\n \"messages\": [\n {\n \"role\": \"user\",\n \"content\": \"What is the weather in Santa Clara, CA?\"\n }\n ],\n \"model\": \"meta/llama-3.3-70b-instruct\",\n \"max_tokens\": 1024,\n \"stream\": true,\n \"tool_choice\": \"auto\",\n \"tools\": [\n {\n \"type\": \"function\",\n \"function\": {\n \"name\": \"get_current_weather\",\n \"description\": \"A tool that gets the current weather at a location, if one is specified, and defaults to the user's location.\",\n \"parameters\": {\n \"type\": \"object\",\n \"properties\": {\n \"location\": {\n \"type\": \"string\",\n \"description\": \"The location to find the weather of, or if not provided, it's the default location.\"\n },\n \"unit\": {\n \"type\": \"string\",\n \"enum\": [\n \"u\",\n \"m\"\n ],\n \"description\": \"Whether to use SI or USCS units (celsius or fahrenheit). Infer this from the user's location.\"\n }\n }\n }\n }\n }\n ]\n}\n22e:{\"name\":\"What is the weather in Santa Clara, CA?\",\"requestJson\":\"$22f\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"obj"])</script><script>self.__next_f.push([1,"ect\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n22a:[\"$22b\",\"$22c\",\"$22e\"]\n233:T4bd,from openai import OpenAI\n\nclient = OpenAI(\n base_url = \"https://integrate.api.nvidia.com/v1\",\n api_key = \"$NVIDIA_API_KEY\"\n)\n\u003c% if (request.tools) { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e,\n tools=\u003c%- JSON.stringify(request.tools) %\u003e,\n \u003c% if (request.tool_choice) { %\u003etool_choice=\u003c%- JSON.stringify(request.tool_choice) %\u003e\u003c% } %\u003e\n)\u003c% } else { %\u003e\ncompletion = client.chat.completions.create(\n model=\"\u003c%- request.model %\u003e\",\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\n temperature=\u003c%- request.temperature %\u003e,\n top_p=\u003c%- request.top_p %\u003e,\n max_tokens=\u003c%- request.max_tokens %\u003e,\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\n)\u003c% } %\u003e\n\u003c% if (request.stream) { %\u003e\nfor chunk in completion:\n if chunk.choices[0].delta.content is not None:\n print(chunk.choices[0].delta.content, end=\"\")\n\u003c% } else { %\u003e\nprint(completion.choices[0].message)\n\u003c% } %\u003e\n234:T504,import OpenAI from 'openai';\n\nconst openai = new OpenAI({\n apiKey: '$NVIDIA_API_KEY',\n baseURL: 'https://integrate.api.nvidia.com/v1',\n})\n \u003c% if (request.tools) { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify"])</script><script>self.__next_f.push([1,"(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e,\n \u003c% if (request.tools) { %\u003etools: \u003c%- JSON.stringify(request.tools) %\u003e,\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003etool_choice: \u003c%- JSON.stringify(request.tool_choice) %\u003e,\u003c% } %\u003e\n })\u003c% } else { %\u003e\nasync function main() {\n const completion = await openai.chat.completions.create({\n model: \"\u003c%- request.model %\u003e\",\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\n temperature: \u003c%- request.temperature %\u003e,\n top_p: \u003c%- request.top_p %\u003e,\n max_tokens: \u003c%- request.max_tokens %\u003e,\n stream: \u003c%- request.stream %\u003e\n })\u003c% } %\u003e\n \u003c% if (request.stream) { %\u003e\n for await (const chunk of completion) {\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\n }\n \u003c% } else { %\u003e\n process.stdout.write(completion.choices[0]?.message?.content);\n \u003c% } %\u003e\n}\n\nmain();235:T66f,\u003c% if (request.tools) { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } else { %\u003e\n \"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n\n -H \\\"Content-Type: application/json\\\" \\\\\\n\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n\n -d '{\\n\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n\n "])</script><script>self.__next_f.push([1," \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n\n \\\"temperature\\\": \u003c%- request.temperature %\u003e,\\n\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n\n \\\"stream\\\": \u003c%- request.stream %\u003e\n \u003c% if (request.tools) { %\u003e,\\n \\\"tools\\\": \u003c%- JSON.stringify(request.tools).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n \u003c% if (request.tool_choice) { %\u003e,\\n \\\"tool_choice\\\": \u003c%- JSON.stringify(request.tool_choice).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\u003c% } %\u003e\n }'\\n\"\u003c% } %\u003e232:{\"python\":\"$233\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n \\n\",\"node.js\":\"$234\",\"curl\":\"$235\"}\n231:{\"title\":\"No Streaming\",\"requestEjs\":\"$232\",\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n230:[\"$231\"]\n229:{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":\"$22a\",\"templates\":\"$230\"}\n20d:{\"op"])</script><script>self.__next_f.push([1,"erationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":\"$20e\",\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":\"$20f\",\"responses\":\"$212\",\"x-nvai-meta\":\"$229\"}\n20c:{\"post\":\"$20d\"}\n20b:{\"/chat/completions\":\"$20c\"}\n238:[]\n237:{\"Token\":\"$238\"}\n236:[\"$237\"]\n23b:{\"type\":\"http\",\"scheme\":\"bearer\"}\n23a:{\"Token\":\"$23b\"}\n240:[\"function\"]\n23f:{\"type\":\"string\",\"enum\":\"$240\",\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"}\n23e:{\"function\":\"$1fb\",\"type\":\"$23f\"}\n241:[\"function\"]\n23d:{\"properties\":\"$23e\",\"additionalProperties\":false,\"type\":\"object\",\"required\":\"$241\",\"title\":\"ChatCompletionNamedToolChoiceParam\"}\n23c:{\"Errors\":\"$194\",\"ChatCompletion\":\"$1a2\",\"ChatCompletionChunk\":\"$1c0\",\"ChatRequest\":\"$1d2\",\"Choice\":\"$1a6\",\"ChoiceChunk\":\"$1c4\",\"Message\":\"$172\",\"ToolCall\":\"$181\",\"FunctionCall\":\"$186\",\"ChatCompletionToolsParam\":\"$1de\",\"ChatCompletionNamedFunction\":\"$1fb\",\"ChatCompletionNamedToolChoiceParam\":\"$23d\",\"FunctionDefinition\":\"$1e2\",\"Usage\":\"$1b6\"}\n239:{\"securitySchemes\":\"$23a\",\"schemas\":\"$23c\"}\n205:{\"openapi\":\"3.1.0\",\"info\":\"$206\",\"servers\":\"$209\",\"paths\":\"$20b\",\"security\":\"$236\",\"components\":\"$239\"}\n242:T57ca,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [ -0.0175628662109375, -0.035247802734375, 0.044586181640625, -0.034881591796875, 0.0277099609375, -0.01152801513671875, 0.00598907470703125, -0.026031494140625, 0.023895263671875, -0.00873565673828125, 0.01508331298828125, 0.052337646484375, -0.00351715087890625, -0.005550384521484375, 0.018768310546875, -0.0264739990234375, -0.0262298583984375, -0.01971435546875, 0.040008544921875, -0.019195556640625, -0.01557159423828125, -0.023468017578125, 0.038665771484375, 0.0242919921875, -0.03631591796875, 0.041290283203125, 0.01129150390625, 0.0255584716796875, 0.007411956787109375, 0.005908966064453125, 0.03338623046875, -0.0628662109375, 0.01464080810546875, 0.046295166015625, 0.0084381103515625, -0.040802001953125, -0.039215087890625, -0.0233612060546875, -0.0482177734375, 0.01471710205078125, -0.0226287841796875, -0.030731201171875, 0.020599365234375, 0.09307861328125, 0.00038886070251464844, -0.00815582275390625, 0.00786590576171875, -0.06182861328125, 0.0242919921875, -0.05206298828125, 0.0254058837890625, 0.0254669189453125, 0.06396484375, 0.0178375244140625, -0.032135009765625, 0.0184326171875, 0.0609130859375, 0.071044921875, -0.0197906494140625, -0.036956787109375, 0.0290985107421875, -0.078857421875, -0.0015840530395507812, -0.0198974609375, -0.033355712890625, 0.0084686279296875, 0.002529144287109375, -0.044403076171875, 0.015869140625, -0.0236358642578125, -0.0005974769592285156, -0.020599365234375, 0.017730712890625, -0.02294921875, 0.043548583984375, -0.050689697265625, -0.035186767578125, 0.02178955078125, 0.013580322265625, 0.0615234375, -0.01495361328125, 0.00872802734375, 0.0018701553344726562, -0.0248260498046875, 0.00992584228515625, -0.01361083984375, 0.0709228515625, 0.05712890625, -0.015869140625, -0.0007290840148925781, -0.033355712890625, -0.0078582763671875, 0.052398681640625, -0.00958251953125, -0.0095062255859375, 0.01354217529296875, -0.0176239013671875, 0.0283966064453125, 0.00710296630859375, -0.0223541259765625, -0.01190185546875, -0.01412200927734375, 0.00897979736328125, 0.037567138671875, 0.0009965896606445312, 0.0201873779296875, 0.03729248046875, 0.0362548828125, -0.03704833984375, 0.02081298828125, -0.0182342529296875, -0.00568389892578125, 0.00821685791015625, 0.01947021484375, 0.00022363662719726562, -0.005825042724609375, -0.020233154296875, -0.0277099609375, -0.013671875, 0.051666259765625, 0.018157958984375, 0.01837158203125, 0.0301055908203125, 0.01177215576171875, -0.04156494140625, 0.06207275390625, -0.016143798828125, -0.031646728515625, -0.061248779296875, -0.0059051513671875, -0.0248260498046875, -0.033599853515625, 0.05169677734375, -0.0244903564453125, 0.0013380050659179688, -0.0263671875, -0.048248291015625, -0.0509033203125, 0.0272369384765625, -0.0011968612670898438, 0.0088348388671875, -0.0264892578125, -0.003879547119140625, 0.00629425048828125, 0.062103271484375, 0.0308380126953125, 0.01055908203125, -0.00960540771484375, 0.00162506103515625, 0.01461029052734375, -0.03765869140625, 0.0278167724609375, -0.0286407470703125, 0.0019464492797851562, 0.0177764892578125, -0.0230865478515625, -0.01122283935546875, 0.060272216796875, -0.027984619140625, -0.0152435302734375, 0.010498046875, 0.0106201171875, 0.0055084228515625, 0.018341064453125, -0.004058837890625, -0.0212554931640625, 0.04119873046875, -0.0213165283203125, 0.05401611328125, -0.006816864013671875, -0.047576904296875, -0.00685882568359375, 0.0654296875, 0.007343292236328125, 0.040435791015625, -0.006717681884765625, 0.0321044921875, -0.013458251953125, -0.026214599609375, 0.051971435546875, -0.035888671875, -0.0175628662109375, 0.043731689453125, 0.0458984375, 0.03125, 0.0158843994140625, -0.03900146484375, 0.01239013671875, -0.005107879638671875, 0.0009927749633789062, -0.063232421875, -0.0579833984375, 0.057342529296875, -0.0162506103515625, 0.0318603515625, 0.019989013671875, 0.0159759521484375, 0.00249481201171875, 0.019012451171875, -0.055450439453125, -0.038726806640625, 0.01552581787109375, 0.0128326416015625, -0.0222930908203125, 0.005161285400390625, 0.0555419921875, 0.010986328125, -0.017059326171875, -0.018829345703125, 0.004001617431640625, -0.041351318359375, 0.00566864013671875, -0.02593994140625, -0.0026683807373046875, -0.064697265625, -0.0137176513671875, 0.00902557373046875, 0.0211181640625, -0.0041656494140625, 0.039886474609375, -0.01611328125, 0.05657958984375, 0.0609130859375, 0.03179931640625, 0.0052947998046875, -0.005023956298828125, -0.061309814453125, -0.015960693359375, 0.002635955810546875, 0.0257720947265625, -0.055450439453125, -0.00621795654296875, 0.047271728515625, 0.000043451786041259766, 0.042022705078125, -0.023681640625, 0.047088623046875, -0.03814697265625, 0.0726318359375, 0.03350830078125, -0.02716064453125, 0.0011434555053710938, 0.0233612060546875, 0.04119873046875, 0.01219940185546875, 0.00318145751953125, 0.0142364501953125, 0.02935791015625, 0.05499267578125, 0.0142822265625, -0.012359619140625, 0.054046630859375, 0.0023365020751953125, -0.06536865234375, -0.005855560302734375, -0.012115478515625, 0.027984619140625, -0.048675537109375, -0.01800537109375, 0.08160400390625, -0.0292816162109375, 0.03192138671875, 0.0171966552734375, 0.019073486328125, 0.0187225341796875, -0.036376953125, 0.02142333984375, -0.048065185546875, 0.062286376953125, -0.051544189453125, 0.00608062744140625, -0.00591278076171875, -0.01114654541015625, -0.0303802490234375, 0.00946044921875, 0.031951904296875, 0.0052032470703125, -0.0245513916015625, -0.03326416015625, -0.02191162109375, 0.03497314453125, -0.0350341796875, 0.0035991668701171875, -0.022186279296875, -0.04864501953125, -0.026214599609375, -0.00824737548828125, 0.00009715557098388672, 0.01511383056640625, -0.02301025390625, 0.00632476806640625, 0.040863037109375, -0.034759521484375, -0.0321044921875, 0.0027599334716796875, 0.00803375244140625, 0.01885986328125, -0.0233001708984375, 0.0233154296875, 0.010955810546875, 0.0154571533203125, -0.057769775390625, 0.0240020751953125, 0.03411865234375, -0.03131103515625, -0.0181732177734375, 0.0028209686279296875, 0.01348114013671875, -0.06884765625, 0.01412200927734375, 0.0288848876953125, -0.02581787109375, 0.042816162109375, 0.0017719268798828125, -0.02093505859375, 0.0259552001953125, -0.023040771484375, 0.011474609375, -0.01812744140625, -0.01099395751953125, 0.03326416015625, 0.01380157470703125, -0.0147705078125, -0.01702880859375, -0.0026912689208984375, 0.01593017578125, -0.0187225341796875, 0.027008056640625, 0.0180206298828125, 0.0164642333984375, -0.036041259765625, -0.01776123046875, -0.01363372802734375, 0.050201416015625, -0.0019388198852539062, -0.030487060546875, 0.0010833740234375, 0.007232666015625, -0.02899169921875, 0.054901123046875, -0.004779815673828125, -0.01276397705078125, 0.035186767578125, -0.023681640625, 0.0029201507568359375, 0.0106201171875, -0.0272674560546875, 0.06707763671875, 0.004215240478515625, 0.04339599609375, -0.01486968994140625, -0.037445068359375, 0.0182952880859375, 0.0102386474609375, -0.01468658447265625, 0.005229949951171875, 0.0017652511596679688, -0.02203369140625, 0.0117340087890625, 0.036895751953125, -0.01132965087890625, -0.0045318603515625, -0.00010925531387329102, -0.031982421875, 0.0010833740234375, 0.02093505859375, -0.02874755859375, -0.05914306640625, -0.055206298828125, 0.075439453125, 0.0081024169921875, 0.0020694732666015625, -0.011077880859375, 0.04119873046875, -0.071044921875, -0.0150146484375, -0.031982421875, 0.009185791015625, 0.029541015625, 0.01136016845703125, 0.04522705078125, 0.01318359375, 0.01824951171875, -0.00812530517578125, -0.03973388671875, -0.0474853515625, -0.0242462158203125, 0.0013980865478515625, 0.01291656494140625, 0.00904083251953125, -0.005100250244140625, 0.01007843017578125, 0.00856781005859375, -0.021484375, -0.011688232421875, -0.045379638671875, 0.0084381103515625, 0.00750732421875, 0.01369476318359375, -0.0264739990234375, 0.0273895263671875, 0.00806427001953125, 0.02557373046875, 0.016571044921875, -0.005931854248046875, -0.028045654296875, 0.0035648345947265625, 0.0477294921875, 0.0216522216796875, -0.052215576171875, 0.0433349609375, 0.0225677490234375, 0.00005179643630981445, -0.0291595458984375, -0.021453857421875, -0.0025234222412109375, 0.006702423095703125, 0.00835418701171875, 0.00324249267578125, 0.039093017578125, 0.0169677734375, -0.02435302734375, -0.000054836273193359375, 0.032623291015625, -0.0325927734375, -0.01000213623046875, 0.051849365234375, 0.039306640625, -0.0281829833984375, 0.03369140625, 0.05340576171875, 0.0165863037109375, -0.00799560546875, -0.08905029296875, -0.02764892578125, -0.0322265625, -0.0279693603515625, -0.0007419586181640625, -0.021484375, -0.004199981689453125, -0.0205535888671875, -0.019500732421875, -0.01027679443359375, -0.024658203125, -0.01139068603515625, 0.0157318115234375, 0.04345703125, -0.01532745361328125, 0.028472900390625, -0.038909912109375, 0.005680084228515625, 0.0246429443359375, -0.01023101806640625, -0.0345458984375, -0.066650390625, 0.006328582763671875, 0.014984130859375, -0.002597808837890625, 0.042694091796875, 0.0108489990234375, -0.0198211669921875, 0.0120849609375, -0.008544921875, -0.045318603515625, 0.005558013916015625, -0.01230621337890625, 0.0253448486328125, -0.044586181640625, -0.0198211669921875, -0.0151519775390625, 0.019439697265625, -0.00640106201171875, 0.0232086181640625, -0.04296875, -0.0191497802734375, 0.002742767333984375, 0.045654296875, -0.0179595947265625, 0.07421875, 0.0421142578125, 0.0227813720703125, -0.01195526123046875, 0.0162506103515625, -0.0121917724609375, -0.0574951171875, 0.051910400390625, -0.0009026527404785156, 0.024658203125, 0.01375579833984375, -0.01454925537109375, 0.0163421630859375, 0.028076171875, 0.0301055908203125, 0.0103607177734375, -0.03955078125, -0.003047943115234375, 0.047943115234375, 0.0169830322265625, 0.03558349609375, 0.036102294921875, 0.037078857421875, 0.023895263671875, -0.01000213623046875, 0.0193939208984375, 0.03814697265625, -0.0648193359375, -0.0247039794921875, -0.0149078369140625, -0.007965087890625, -0.044952392578125, 0.0248260498046875, 0.00893402099609375, 0.0302734375, -0.002170562744140625, -0.0293121337890625, 0.01580810546875, 0.03387451171875, 0.032257080078125, -0.037689208984375, -0.0086517333984375, -0.0103912353515625, -0.0379638671875, 0.0018854141235351562, -0.027191162109375, -0.07049560546875, 0.0009918212890625, 0.031494140625, 0.01355743408203125, 0.01190185546875, 0.006916046142578125, 0.024566650390625, -0.0012769699096679688, -0.01203155517578125, 0.032440185546875, -0.034210205078125, 0.03302001953125, -0.02581787109375, -0.041900634765625, 0.00798797607421875, 0.0013275146484375, -0.0472412109375, 0.033843994140625, -0.04864501953125, 0.01702880859375, -0.0016794204711914062, 0.0253448486328125, -0.0789794921875, -0.055450439453125, 0.00783538818359375, 0.004093170166015625, 0.0233001708984375, -0.052001953125, 0.0201416015625, -0.052337646484375, 0.005352020263671875, 0.031646728515625, 0.0277252197265625, 0.04071044921875, 0.01430511474609375, 0.009490966796875, -0.0421142578125, 0.07415771484375, 0.0382080078125, -0.019744873046875, 0.004650115966796875, 0.0178680419921875, -0.0268707275390625, -0.0109710693359375, 0.004146575927734375, -0.00307464599609375, 0.0546875, -0.028411865234375, -0.03424072265625, 0.038177490234375, -0.0372314453125, 0.03271484375, 0.003421783447265625, -0.007236480712890625, 0.0210113525390625, -0.02947998046875, 0.017730712890625, -0.05706787109375, 0.034027099609375, 0.006542205810546875, -0.007518768310546875, -0.006488800048828125, 0.034332275390625, 0.024566650390625, 0.0264739990234375, 0.0074920654296875, -0.006229400634765625, 0.06500244140625, -0.0206451416015625, -0.009063720703125, -0.012481689453125, -0.044281005859375, 0.01044464111328125, -0.0023441314697265625, 0.045379638671875, 0.003055572509765625, -0.0290374755859375, -0.0233154296875, -0.0345458984375, -0.05059814453125, 0.04388427734375, 0.0146026611328125, 0.0113067626953125, -0.01285552978515625, -0.00010091066360473633, -0.02874755859375, 0.013275146484375, 0.000316619873046875, 0.0192413330078125, 0.0279083251953125, -0.024993896484375, 0.0362548828125, -0.01552581787109375, -0.0137176513671875, 0.044158935546875, 0.06707763671875, -0.0111541748046875, 0.01326751708984375, -0.004917144775390625, -0.0078582763671875, 0.057647705078125, -0.0130615234375, -0.030487060546875, 0.00970458984375, 0.0168304443359375, 0.032501220703125, -0.033050537109375, -0.049530029296875, 0.020172119140625, -0.0255584716796875, 0.03271484375, -0.01384735107421875, -0.00930023193359375, -0.028961181640625, -0.0292816162109375, 0.02642822265625, -0.01337432861328125, -0.003292083740234375, 0.045684814453125, -0.00431060791015625, 0.0499267578125, 0.036407470703125, -0.00847625732421875, -0.051910400390625, -0.051239013671875, 0.033447265625, -0.0272979736328125, 0.03948974609375, 0.0672607421875, -0.0183258056640625, 0.0433349609375, 0.034881591796875, 0.0338134765625, -0.036102294921875, 0.01422882080078125, 0.02996826171875, 0.079345703125, -0.00345611572265625, 0.020263671875, -0.0188751220703125, 0.0024471282958984375, 0.12042236328125, 0.052886962890625, 0.0256805419921875, -0.044464111328125, -0.05120849609375, -0.0279388427734375, 0.0195770263671875, -0.005786895751953125, -0.01042938232421875, 0.00942230224609375, -0.0123748779296875, 0.061676025390625, -0.00473785400390625, 0.01107025146484375, 0.0299072265625, -0.0079803466796875, -0.003692626953125, 0.08154296875, 0.0176849365234375, -0.00885009765625, 0.019683837890625, -0.01186370849609375, -0.06512451171875, -0.017822265625, 0.0162811279296875, -0.0219879150390625, 0.023529052734375, -0.0206451416015625, -0.01546478271484375, 0.023529052734375, -0.003368377685546875, 0.016876220703125, -0.004364013671875, 0.0194091796875, 0.021697998046875, -0.0034313201904296875, 0.0009307861328125, 0.062744140625, -0.037353515625, 0.004169464111328125, 0.0147857666015625, 0.00099945068359375, -0.0340576171875, 0.0216522216796875, -0.0211029052734375, 0.00015842914581298828, 0.01358795166015625, -0.00841522216796875, -0.033203125, -0.01336669921875, 0.06884765625, -0.0025482177734375, -0.0120697021484375, -0.030975341796875, 0.0457763671875, -0.0087738037109375, 0.0220947265625, -0.017822265625, 0.0025806427001953125, -0.017852783203125, 0.0217742919921875, -0.059906005859375, 0.02294921875, 0.0033359527587890625, -0.035125732421875, 0.01311492919921875, -0.0094757080078125, 0.0226287841796875, -0.03997802734375, -0.05792236328125, 0.0246734619140625, -0.01450347900390625, -0.019256591796875, -0.04718017578125, 0.031890869140625, -0.018157958984375, -0.00597381591796875, 0.004894256591796875, -0.05950927734375, -0.0204010009765625, -0.038360595703125, 0.001361846923828125, 0.0255584716796875, -0.01404571533203125, -0.0295257568359375, -0.0223388671875, 0.06024169921875, -0.004787445068359375, -0.047607421875, 0.0157012939453125, -0.032806396484375, 0.00218963623046875, 0.0152587890625, -0.040435791015625, -0.017791748046875, 0.050384521484375, -0.0592041015625, 0.0002696514129638672, 0.01259613037109375, 0.04437255859375, 0.0333251953125, 0.02099609375, 0.00826263427734375, 0.0236358642578125, 0.0389404296875, -0.0009889602661132812, -0.01366424560546875, -0.03240966796875, -0.007049560546875, 0.019500732421875, 0.0249176025390625, -0.00579833984375, -0.0262603759765625, -0.07281494140625, -0.051666259765625, 0.037628173828125, -0.01885986328125, -0.0394287109375, 0.00856781005859375, 0.024993896484375, 0.008880615234375, 0.019287109375, 0.06622314453125, 0.023223876953125, 0.043731689453125, 0.06640625, 0.012298583984375, 0.0595703125, -0.02691650390625, -0.02642822265625, 0.036102294921875, 0.037445068359375, 0.0033168792724609375, -0.01294708251953125, 0.00689697265625, 0.005443572998046875, 0.01410675048828125, 0.045867919921875, -0.050628662109375, -0.014984130859375, 0.0390625, -0.070556640625, -0.0242156982421875, 0.007061004638671875, -0.0152587890625, -0.00603485107421875, 0.0004017353057861328, -0.0261688232421875, 0.032928466796875, 0.050567626953125, -0.0177001953125, -0.0259552001953125, 0.0263214111328125, -0.040130615234375, 0.031982421875, 0.028717041015625, 0.0164337158203125, -0.01131439208984375, 0.0271759033203125, 0.00821685791015625, 0.024383544921875, -0.052734375, 0.06640625, -0.0063018798828125, -0.016143798828125, -0.0223541259765625, 0.0189056396484375, -0.042266845703125, -0.00015866756439208984, 0.007030487060546875, 0.0294342041015625, 0.0091705322265625, -0.030914306640625, 0.0313720703125, -0.05584716796875, 0.01409149169921875, -0.0035228729248046875, -0.0231475830078125, 0.023101806640625, 0.01361083984375, 0.02545166015625, 0.052825927734375, -0.0521240234375, 0.0298004150390625, 0.004116058349609375, -0.041259765625, -0.04693603515625, -0.002010345458984375, 0.0311431884765625, -0.0270538330078125, 0.03143310546875, 0.0022449493408203125, 0.0207977294921875, -0.0002307891845703125, -0.037017822265625, -0.0272216796875, 0.0009908676147460938, 0.033538818359375, -0.053253173828125, -0.01450347900390625, -0.00408935546875, 0.00792694091796875, -0.003879547119140625, 0.02825927734375, 0.0291748046875, -0.01529693603515625, 0.01099395751953125, 0.0889892578125, 0.03289794921875, 0.0460205078125, 0.0306549072265625, -0.031982421875, -0.036651611328125, -0.04254150390625, -0.0194091796875, 0.03759765625, 0.00347900390625, 0.06036376953125, -0.025421142578125, -0.00423431396484375, -0.003814697265625, 0.02587890625, -0.041900634765625, -0.0789794921875, -0.01227569580078125, -0.0214996337890625, 0.037445068359375, -0.02197265625, -0.019256591796875, -0.01383209228515625, -0.026947021484375, -0.00579833984375, -0.03240966796875, -0.050201416015625, 0.006320953369140625, -0.02374267578125, -0.026702880859375, -0.004070281982421875, -0.01287078857421875, -0.0229034423828125, -0.04022216796875, -0.035888671875, -0.020172119140625, 0.00264739990234375, 0.005374908447265625, -0.029815673828125, -0.033905029296875, 0.059844970703125, 0.0062713623046875, 0.041900634765625, 0.04034423828125, 0.00513458251953125, 0.0091705322265625, -0.04864501953125, -0.00991058349609375, -7.152557373046875e-7, 0.000392913818359375, -0.00359344482421875, 0.00519561767578125, -0.016815185546875, -0.00007575750350952148, 0.03521728515625, -0.0215911865234375, -0.023162841796875, -0.0240325927734375, -0.018951416015625, 0.030975341796875, -0.01117706298828125, -0.008331298828125, -0.01529693603515625, -0.05206298828125, 0.0012445449829101562, 0.0023136138916015625, 0.064208984375, -0.007354736328125, -0.03753662109375, 0.0007100105285644531, 0.0036830902099609375, 0.00608062744140625, -0.01541900634765625, -0.021087646484375, 0.0225067138671875, 0.00518035888671875, 0.05523681640625, 0.00641632080078125, -0.0830078125, 0.061248779296875, 0.045318603515625, -0.064208984375, -0.023193359375, 0.0083770751953125, -0.004787445068359375, 0.04266357421875, 0.0689697265625, 0.0213470458984375, -0.06317138671875, 0.028350830078125, 0.0179290771484375, -0.03997802734375, -0.00140380859375, 0.031646728515625, -0.019287109375, -0.0117950439453125, 0.0109100341796875, 0.01340484619140625, -0.046295166015625, -0.03973388671875, 0.01273345947265625, 0.036468505859375, -0.01299285888671875, -0.00962066650390625, 0.032135009765625, 0.0130157470703125, 0.00604248046875, 0.0191192626953125, -0.0135650634765625, -0.003337860107421875, 0.0054779052734375, -0.058868408203125, -0.00634002685546875, -0.0305938720703125, -0.019989013671875, 0.024444580078125, 0.0097503662109375, 0.07818603515625, -0.03997802734375, 0.0227813720703125, -0.0131683349609375, -0.007534027099609375, -0.045257568359375, 0.0631103515625, 0.02471923828125, 0.038665771484375, 0.01050567626953125, 0.0038280487060546875, 0.046875, -0.033203125, 0.035308837890625, -0.037872314453125, -0.0115203857421875, -0.039520263671875, -0.0279693603515625, -0.01386260986328125, -0.0177001953125, 0.039520263671875, -0.016387939453125, -0.03021240234375, 0.0186920166015625, -0.008392333984375, 0.01145172119140625, -0.0269775390625, -0.016082763671875, 0.0242919921875, -0.039794921875, 0.0809326171875, -0.01479339599609375, -0.039520263671875, -0.00608062744140625, 0.030303955078125],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/embed-qa-4\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"243:T57ca,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [ -0.0175628662109375, -0.035247802734375, 0.044586181640625, -0.034881591796875, 0.0277099609375, -0.01152801513671875, 0.00598907470703125, -0.026031494140625, 0.023895263671875, -0.00873565673828125, 0.01508331298828125, 0.052337646484375, -0.00351715087890625, -0.005550384521484375, 0.018768310546875, -0.0264739990234375, -0.0262298583984375, -0.01971435546875, 0.040008544921875, -0.019195556640625, -0.01557159423828125, -0.023468017578125, 0.038665771484375, 0.0242919921875, -0.03631591796875, 0.041290283203125, 0.01129150390625, 0.0255584716796875, 0.007411956787109375, 0.005908966064453125, 0.03338623046875, -0.0628662109375, 0.01464080810546875, 0.046295166015625, 0.0084381103515625, -0.040802001953125, -0.039215087890625, -0.0233612060546875, -0.0482177734375, 0.01471710205078125, -0.0226287841796875, -0.030731201171875, 0.020599365234375, 0.09307861328125, 0.00038886070251464844, -0.00815582275390625, 0.00786590576171875, -0.06182861328125, 0.0242919921875, -0.05206298828125, 0.0254058837890625, 0.0254669189453125, 0.06396484375, 0.0178375244140625, -0.032135009765625, 0.0184326171875, 0.0609130859375, 0.071044921875, -0.0197906494140625, -0.036956787109375, 0.0290985107421875, -0.078857421875, -0.0015840530395507812, -0.0198974609375, -0.033355712890625, 0.0084686279296875, 0.002529144287109375, -0.044403076171875, 0.015869140625, -0.0236358642578125, -0.0005974769592285156, -0.020599365234375, 0.017730712890625, -0.02294921875, 0.043548583984375, -0.050689697265625, -0.035186767578125, 0.02178955078125, 0.013580322265625, 0.0615234375, -0.01495361328125, 0.00872802734375, 0.0018701553344726562, -0.0248260498046875, 0.00992584228515625, -0.01361083984375, 0.0709228515625, 0.05712890625, -0.015869140625, -0.0007290840148925781, -0.033355712890625, -0.0078582763671875, 0.052398681640625, -0.00958251953125, -0.0095062255859375, 0.01354217529296875, -0.0176239013671875, 0.0283966064453125, 0.00710296630859375, -0.0223541259765625, -0.01190185546875, -0.01412200927734375, 0.00897979736328125, 0.037567138671875, 0.0009965896606445312, 0.0201873779296875, 0.03729248046875, 0.0362548828125, -0.03704833984375, 0.02081298828125, -0.0182342529296875, -0.00568389892578125, 0.00821685791015625, 0.01947021484375, 0.00022363662719726562, -0.005825042724609375, -0.020233154296875, -0.0277099609375, -0.013671875, 0.051666259765625, 0.018157958984375, 0.01837158203125, 0.0301055908203125, 0.01177215576171875, -0.04156494140625, 0.06207275390625, -0.016143798828125, -0.031646728515625, -0.061248779296875, -0.0059051513671875, -0.0248260498046875, -0.033599853515625, 0.05169677734375, -0.0244903564453125, 0.0013380050659179688, -0.0263671875, -0.048248291015625, -0.0509033203125, 0.0272369384765625, -0.0011968612670898438, 0.0088348388671875, -0.0264892578125, -0.003879547119140625, 0.00629425048828125, 0.062103271484375, 0.0308380126953125, 0.01055908203125, -0.00960540771484375, 0.00162506103515625, 0.01461029052734375, -0.03765869140625, 0.0278167724609375, -0.0286407470703125, 0.0019464492797851562, 0.0177764892578125, -0.0230865478515625, -0.01122283935546875, 0.060272216796875, -0.027984619140625, -0.0152435302734375, 0.010498046875, 0.0106201171875, 0.0055084228515625, 0.018341064453125, -0.004058837890625, -0.0212554931640625, 0.04119873046875, -0.0213165283203125, 0.05401611328125, -0.006816864013671875, -0.047576904296875, -0.00685882568359375, 0.0654296875, 0.007343292236328125, 0.040435791015625, -0.006717681884765625, 0.0321044921875, -0.013458251953125, -0.026214599609375, 0.051971435546875, -0.035888671875, -0.0175628662109375, 0.043731689453125, 0.0458984375, 0.03125, 0.0158843994140625, -0.03900146484375, 0.01239013671875, -0.005107879638671875, 0.0009927749633789062, -0.063232421875, -0.0579833984375, 0.057342529296875, -0.0162506103515625, 0.0318603515625, 0.019989013671875, 0.0159759521484375, 0.00249481201171875, 0.019012451171875, -0.055450439453125, -0.038726806640625, 0.01552581787109375, 0.0128326416015625, -0.0222930908203125, 0.005161285400390625, 0.0555419921875, 0.010986328125, -0.017059326171875, -0.018829345703125, 0.004001617431640625, -0.041351318359375, 0.00566864013671875, -0.02593994140625, -0.0026683807373046875, -0.064697265625, -0.0137176513671875, 0.00902557373046875, 0.0211181640625, -0.0041656494140625, 0.039886474609375, -0.01611328125, 0.05657958984375, 0.0609130859375, 0.03179931640625, 0.0052947998046875, -0.005023956298828125, -0.061309814453125, -0.015960693359375, 0.002635955810546875, 0.0257720947265625, -0.055450439453125, -0.00621795654296875, 0.047271728515625, 0.000043451786041259766, 0.042022705078125, -0.023681640625, 0.047088623046875, -0.03814697265625, 0.0726318359375, 0.03350830078125, -0.02716064453125, 0.0011434555053710938, 0.0233612060546875, 0.04119873046875, 0.01219940185546875, 0.00318145751953125, 0.0142364501953125, 0.02935791015625, 0.05499267578125, 0.0142822265625, -0.012359619140625, 0.054046630859375, 0.0023365020751953125, -0.06536865234375, -0.005855560302734375, -0.012115478515625, 0.027984619140625, -0.048675537109375, -0.01800537109375, 0.08160400390625, -0.0292816162109375, 0.03192138671875, 0.0171966552734375, 0.019073486328125, 0.0187225341796875, -0.036376953125, 0.02142333984375, -0.048065185546875, 0.062286376953125, -0.051544189453125, 0.00608062744140625, -0.00591278076171875, -0.01114654541015625, -0.0303802490234375, 0.00946044921875, 0.031951904296875, 0.0052032470703125, -0.0245513916015625, -0.03326416015625, -0.02191162109375, 0.03497314453125, -0.0350341796875, 0.0035991668701171875, -0.022186279296875, -0.04864501953125, -0.026214599609375, -0.00824737548828125, 0.00009715557098388672, 0.01511383056640625, -0.02301025390625, 0.00632476806640625, 0.040863037109375, -0.034759521484375, -0.0321044921875, 0.0027599334716796875, 0.00803375244140625, 0.01885986328125, -0.0233001708984375, 0.0233154296875, 0.010955810546875, 0.0154571533203125, -0.057769775390625, 0.0240020751953125, 0.03411865234375, -0.03131103515625, -0.0181732177734375, 0.0028209686279296875, 0.01348114013671875, -0.06884765625, 0.01412200927734375, 0.0288848876953125, -0.02581787109375, 0.042816162109375, 0.0017719268798828125, -0.02093505859375, 0.0259552001953125, -0.023040771484375, 0.011474609375, -0.01812744140625, -0.01099395751953125, 0.03326416015625, 0.01380157470703125, -0.0147705078125, -0.01702880859375, -0.0026912689208984375, 0.01593017578125, -0.0187225341796875, 0.027008056640625, 0.0180206298828125, 0.0164642333984375, -0.036041259765625, -0.01776123046875, -0.01363372802734375, 0.050201416015625, -0.0019388198852539062, -0.030487060546875, 0.0010833740234375, 0.007232666015625, -0.02899169921875, 0.054901123046875, -0.004779815673828125, -0.01276397705078125, 0.035186767578125, -0.023681640625, 0.0029201507568359375, 0.0106201171875, -0.0272674560546875, 0.06707763671875, 0.004215240478515625, 0.04339599609375, -0.01486968994140625, -0.037445068359375, 0.0182952880859375, 0.0102386474609375, -0.01468658447265625, 0.005229949951171875, 0.0017652511596679688, -0.02203369140625, 0.0117340087890625, 0.036895751953125, -0.01132965087890625, -0.0045318603515625, -0.00010925531387329102, -0.031982421875, 0.0010833740234375, 0.02093505859375, -0.02874755859375, -0.05914306640625, -0.055206298828125, 0.075439453125, 0.0081024169921875, 0.0020694732666015625, -0.011077880859375, 0.04119873046875, -0.071044921875, -0.0150146484375, -0.031982421875, 0.009185791015625, 0.029541015625, 0.01136016845703125, 0.04522705078125, 0.01318359375, 0.01824951171875, -0.00812530517578125, -0.03973388671875, -0.0474853515625, -0.0242462158203125, 0.0013980865478515625, 0.01291656494140625, 0.00904083251953125, -0.005100250244140625, 0.01007843017578125, 0.00856781005859375, -0.021484375, -0.011688232421875, -0.045379638671875, 0.0084381103515625, 0.00750732421875, 0.01369476318359375, -0.0264739990234375, 0.0273895263671875, 0.00806427001953125, 0.02557373046875, 0.016571044921875, -0.005931854248046875, -0.028045654296875, 0.0035648345947265625, 0.0477294921875, 0.0216522216796875, -0.052215576171875, 0.0433349609375, 0.0225677490234375, 0.00005179643630981445, -0.0291595458984375, -0.021453857421875, -0.0025234222412109375, 0.006702423095703125, 0.00835418701171875, 0.00324249267578125, 0.039093017578125, 0.0169677734375, -0.02435302734375, -0.000054836273193359375, 0.032623291015625, -0.0325927734375, -0.01000213623046875, 0.051849365234375, 0.039306640625, -0.0281829833984375, 0.03369140625, 0.05340576171875, 0.0165863037109375, -0.00799560546875, -0.08905029296875, -0.02764892578125, -0.0322265625, -0.0279693603515625, -0.0007419586181640625, -0.021484375, -0.004199981689453125, -0.0205535888671875, -0.019500732421875, -0.01027679443359375, -0.024658203125, -0.01139068603515625, 0.0157318115234375, 0.04345703125, -0.01532745361328125, 0.028472900390625, -0.038909912109375, 0.005680084228515625, 0.0246429443359375, -0.01023101806640625, -0.0345458984375, -0.066650390625, 0.006328582763671875, 0.014984130859375, -0.002597808837890625, 0.042694091796875, 0.0108489990234375, -0.0198211669921875, 0.0120849609375, -0.008544921875, -0.045318603515625, 0.005558013916015625, -0.01230621337890625, 0.0253448486328125, -0.044586181640625, -0.0198211669921875, -0.0151519775390625, 0.019439697265625, -0.00640106201171875, 0.0232086181640625, -0.04296875, -0.0191497802734375, 0.002742767333984375, 0.045654296875, -0.0179595947265625, 0.07421875, 0.0421142578125, 0.0227813720703125, -0.01195526123046875, 0.0162506103515625, -0.0121917724609375, -0.0574951171875, 0.051910400390625, -0.0009026527404785156, 0.024658203125, 0.01375579833984375, -0.01454925537109375, 0.0163421630859375, 0.028076171875, 0.0301055908203125, 0.0103607177734375, -0.03955078125, -0.003047943115234375, 0.047943115234375, 0.0169830322265625, 0.03558349609375, 0.036102294921875, 0.037078857421875, 0.023895263671875, -0.01000213623046875, 0.0193939208984375, 0.03814697265625, -0.0648193359375, -0.0247039794921875, -0.0149078369140625, -0.007965087890625, -0.044952392578125, 0.0248260498046875, 0.00893402099609375, 0.0302734375, -0.002170562744140625, -0.0293121337890625, 0.01580810546875, 0.03387451171875, 0.032257080078125, -0.037689208984375, -0.0086517333984375, -0.0103912353515625, -0.0379638671875, 0.0018854141235351562, -0.027191162109375, -0.07049560546875, 0.0009918212890625, 0.031494140625, 0.01355743408203125, 0.01190185546875, 0.006916046142578125, 0.024566650390625, -0.0012769699096679688, -0.01203155517578125, 0.032440185546875, -0.034210205078125, 0.03302001953125, -0.02581787109375, -0.041900634765625, 0.00798797607421875, 0.0013275146484375, -0.0472412109375, 0.033843994140625, -0.04864501953125, 0.01702880859375, -0.0016794204711914062, 0.0253448486328125, -0.0789794921875, -0.055450439453125, 0.00783538818359375, 0.004093170166015625, 0.0233001708984375, -0.052001953125, 0.0201416015625, -0.052337646484375, 0.005352020263671875, 0.031646728515625, 0.0277252197265625, 0.04071044921875, 0.01430511474609375, 0.009490966796875, -0.0421142578125, 0.07415771484375, 0.0382080078125, -0.019744873046875, 0.004650115966796875, 0.0178680419921875, -0.0268707275390625, -0.0109710693359375, 0.004146575927734375, -0.00307464599609375, 0.0546875, -0.028411865234375, -0.03424072265625, 0.038177490234375, -0.0372314453125, 0.03271484375, 0.003421783447265625, -0.007236480712890625, 0.0210113525390625, -0.02947998046875, 0.017730712890625, -0.05706787109375, 0.034027099609375, 0.006542205810546875, -0.007518768310546875, -0.006488800048828125, 0.034332275390625, 0.024566650390625, 0.0264739990234375, 0.0074920654296875, -0.006229400634765625, 0.06500244140625, -0.0206451416015625, -0.009063720703125, -0.012481689453125, -0.044281005859375, 0.01044464111328125, -0.0023441314697265625, 0.045379638671875, 0.003055572509765625, -0.0290374755859375, -0.0233154296875, -0.0345458984375, -0.05059814453125, 0.04388427734375, 0.0146026611328125, 0.0113067626953125, -0.01285552978515625, -0.00010091066360473633, -0.02874755859375, 0.013275146484375, 0.000316619873046875, 0.0192413330078125, 0.0279083251953125, -0.024993896484375, 0.0362548828125, -0.01552581787109375, -0.0137176513671875, 0.044158935546875, 0.06707763671875, -0.0111541748046875, 0.01326751708984375, -0.004917144775390625, -0.0078582763671875, 0.057647705078125, -0.0130615234375, -0.030487060546875, 0.00970458984375, 0.0168304443359375, 0.032501220703125, -0.033050537109375, -0.049530029296875, 0.020172119140625, -0.0255584716796875, 0.03271484375, -0.01384735107421875, -0.00930023193359375, -0.028961181640625, -0.0292816162109375, 0.02642822265625, -0.01337432861328125, -0.003292083740234375, 0.045684814453125, -0.00431060791015625, 0.0499267578125, 0.036407470703125, -0.00847625732421875, -0.051910400390625, -0.051239013671875, 0.033447265625, -0.0272979736328125, 0.03948974609375, 0.0672607421875, -0.0183258056640625, 0.0433349609375, 0.034881591796875, 0.0338134765625, -0.036102294921875, 0.01422882080078125, 0.02996826171875, 0.079345703125, -0.00345611572265625, 0.020263671875, -0.0188751220703125, 0.0024471282958984375, 0.12042236328125, 0.052886962890625, 0.0256805419921875, -0.044464111328125, -0.05120849609375, -0.0279388427734375, 0.0195770263671875, -0.005786895751953125, -0.01042938232421875, 0.00942230224609375, -0.0123748779296875, 0.061676025390625, -0.00473785400390625, 0.01107025146484375, 0.0299072265625, -0.0079803466796875, -0.003692626953125, 0.08154296875, 0.0176849365234375, -0.00885009765625, 0.019683837890625, -0.01186370849609375, -0.06512451171875, -0.017822265625, 0.0162811279296875, -0.0219879150390625, 0.023529052734375, -0.0206451416015625, -0.01546478271484375, 0.023529052734375, -0.003368377685546875, 0.016876220703125, -0.004364013671875, 0.0194091796875, 0.021697998046875, -0.0034313201904296875, 0.0009307861328125, 0.062744140625, -0.037353515625, 0.004169464111328125, 0.0147857666015625, 0.00099945068359375, -0.0340576171875, 0.0216522216796875, -0.0211029052734375, 0.00015842914581298828, 0.01358795166015625, -0.00841522216796875, -0.033203125, -0.01336669921875, 0.06884765625, -0.0025482177734375, -0.0120697021484375, -0.030975341796875, 0.0457763671875, -0.0087738037109375, 0.0220947265625, -0.017822265625, 0.0025806427001953125, -0.017852783203125, 0.0217742919921875, -0.059906005859375, 0.02294921875, 0.0033359527587890625, -0.035125732421875, 0.01311492919921875, -0.0094757080078125, 0.0226287841796875, -0.03997802734375, -0.05792236328125, 0.0246734619140625, -0.01450347900390625, -0.019256591796875, -0.04718017578125, 0.031890869140625, -0.018157958984375, -0.00597381591796875, 0.004894256591796875, -0.05950927734375, -0.0204010009765625, -0.038360595703125, 0.001361846923828125, 0.0255584716796875, -0.01404571533203125, -0.0295257568359375, -0.0223388671875, 0.06024169921875, -0.004787445068359375, -0.047607421875, 0.0157012939453125, -0.032806396484375, 0.00218963623046875, 0.0152587890625, -0.040435791015625, -0.017791748046875, 0.050384521484375, -0.0592041015625, 0.0002696514129638672, 0.01259613037109375, 0.04437255859375, 0.0333251953125, 0.02099609375, 0.00826263427734375, 0.0236358642578125, 0.0389404296875, -0.0009889602661132812, -0.01366424560546875, -0.03240966796875, -0.007049560546875, 0.019500732421875, 0.0249176025390625, -0.00579833984375, -0.0262603759765625, -0.07281494140625, -0.051666259765625, 0.037628173828125, -0.01885986328125, -0.0394287109375, 0.00856781005859375, 0.024993896484375, 0.008880615234375, 0.019287109375, 0.06622314453125, 0.023223876953125, 0.043731689453125, 0.06640625, 0.012298583984375, 0.0595703125, -0.02691650390625, -0.02642822265625, 0.036102294921875, 0.037445068359375, 0.0033168792724609375, -0.01294708251953125, 0.00689697265625, 0.005443572998046875, 0.01410675048828125, 0.045867919921875, -0.050628662109375, -0.014984130859375, 0.0390625, -0.070556640625, -0.0242156982421875, 0.007061004638671875, -0.0152587890625, -0.00603485107421875, 0.0004017353057861328, -0.0261688232421875, 0.032928466796875, 0.050567626953125, -0.0177001953125, -0.0259552001953125, 0.0263214111328125, -0.040130615234375, 0.031982421875, 0.028717041015625, 0.0164337158203125, -0.01131439208984375, 0.0271759033203125, 0.00821685791015625, 0.024383544921875, -0.052734375, 0.06640625, -0.0063018798828125, -0.016143798828125, -0.0223541259765625, 0.0189056396484375, -0.042266845703125, -0.00015866756439208984, 0.007030487060546875, 0.0294342041015625, 0.0091705322265625, -0.030914306640625, 0.0313720703125, -0.05584716796875, 0.01409149169921875, -0.0035228729248046875, -0.0231475830078125, 0.023101806640625, 0.01361083984375, 0.02545166015625, 0.052825927734375, -0.0521240234375, 0.0298004150390625, 0.004116058349609375, -0.041259765625, -0.04693603515625, -0.002010345458984375, 0.0311431884765625, -0.0270538330078125, 0.03143310546875, 0.0022449493408203125, 0.0207977294921875, -0.0002307891845703125, -0.037017822265625, -0.0272216796875, 0.0009908676147460938, 0.033538818359375, -0.053253173828125, -0.01450347900390625, -0.00408935546875, 0.00792694091796875, -0.003879547119140625, 0.02825927734375, 0.0291748046875, -0.01529693603515625, 0.01099395751953125, 0.0889892578125, 0.03289794921875, 0.0460205078125, 0.0306549072265625, -0.031982421875, -0.036651611328125, -0.04254150390625, -0.0194091796875, 0.03759765625, 0.00347900390625, 0.06036376953125, -0.025421142578125, -0.00423431396484375, -0.003814697265625, 0.02587890625, -0.041900634765625, -0.0789794921875, -0.01227569580078125, -0.0214996337890625, 0.037445068359375, -0.02197265625, -0.019256591796875, -0.01383209228515625, -0.026947021484375, -0.00579833984375, -0.03240966796875, -0.050201416015625, 0.006320953369140625, -0.02374267578125, -0.026702880859375, -0.004070281982421875, -0.01287078857421875, -0.0229034423828125, -0.04022216796875, -0.035888671875, -0.020172119140625, 0.00264739990234375, 0.005374908447265625, -0.029815673828125, -0.033905029296875, 0.059844970703125, 0.0062713623046875, 0.041900634765625, 0.04034423828125, 0.00513458251953125, 0.0091705322265625, -0.04864501953125, -0.00991058349609375, -7.152557373046875e-7, 0.000392913818359375, -0.00359344482421875, 0.00519561767578125, -0.016815185546875, -0.00007575750350952148, 0.03521728515625, -0.0215911865234375, -0.023162841796875, -0.0240325927734375, -0.018951416015625, 0.030975341796875, -0.01117706298828125, -0.008331298828125, -0.01529693603515625, -0.05206298828125, 0.0012445449829101562, 0.0023136138916015625, 0.064208984375, -0.007354736328125, -0.03753662109375, 0.0007100105285644531, 0.0036830902099609375, 0.00608062744140625, -0.01541900634765625, -0.021087646484375, 0.0225067138671875, 0.00518035888671875, 0.05523681640625, 0.00641632080078125, -0.0830078125, 0.061248779296875, 0.045318603515625, -0.064208984375, -0.023193359375, 0.0083770751953125, -0.004787445068359375, 0.04266357421875, 0.0689697265625, 0.0213470458984375, -0.06317138671875, 0.028350830078125, 0.0179290771484375, -0.03997802734375, -0.00140380859375, 0.031646728515625, -0.019287109375, -0.0117950439453125, 0.0109100341796875, 0.01340484619140625, -0.046295166015625, -0.03973388671875, 0.01273345947265625, 0.036468505859375, -0.01299285888671875, -0.00962066650390625, 0.032135009765625, 0.0130157470703125, 0.00604248046875, 0.0191192626953125, -0.0135650634765625, -0.003337860107421875, 0.0054779052734375, -0.058868408203125, -0.00634002685546875, -0.0305938720703125, -0.019989013671875, 0.024444580078125, 0.0097503662109375, 0.07818603515625, -0.03997802734375, 0.0227813720703125, -0.0131683349609375, -0.007534027099609375, -0.045257568359375, 0.0631103515625, 0.02471923828125, 0.038665771484375, 0.01050567626953125, 0.0038280487060546875, 0.046875, -0.033203125, 0.035308837890625, -0.037872314453125, -0.0115203857421875, -0.039520263671875, -0.0279693603515625, -0.01386260986328125, -0.0177001953125, 0.039520263671875, -0.016387939453125, -0.03021240234375, 0.0186920166015625, -0.008392333984375, 0.01145172119140625, -0.0269775390625, -0.016082763671875, 0.0242919921875, -0.039794921875, 0.0809326171875, -0.01479339599609375, -0.039520263671875, -0.00608062744140625, 0.030303955078125],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/embed-qa-4\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"246:{\"type\":\"string\",\"example\":\"embedding\"}\n24a:{\"type\":\"number\"}\n249:{\"type\":\"array\",\"items\":\"$24a\",\"description\":\"The embedding vector as a list of floats. The length of the vector depends on the model.\"}\n24b:{\"type\":\"string\",\"description\":\"The embedding vector as a Base64 string. The length of the string depends on the model.\"}\n248:[\"$249\",\"$24b\"]\n247:{\"oneOf\":\"$248\"}\n24c:{\"type\":\"integer\",\"description\":\"The index of the embedding in the list of embeddings.\"}\n245:{\"object\":\"$246\",\"embedding\":\"$247\",\"index\":\"$24c\"}\n24d:[\"object\",\"embedding\",\"index\"]\n244:{\"type\":\"object\",\"properties\":\"$245\",\"required\":\"$24d\"}\n24e:T2260,"])</script><script>self.__next_f.push([1,"## Model Overview\n\n### Description\n\nThe NVIDIA Retrieval QA Embedding Model is an embedding model optimized for text question-answering retrieval. An embedding model is a crucial component of a text retrieval system, as it transforms textual information into dense vector representations. They are typically transformer encoders that process tokens of input text (for example, question, passage) to output an embedding.\n\nNVIDIA Retrieval QA Embedding Model is a part of NVIDIA NeMo Retriever, which provides state-of-the-art, commercially-ready models and microservices, optimized for the lowest latency and highest throughput. It features a production-ready information retrieval pipeline with enterprise support. The models that form the core of this solution have been trained using responsibly selected, auditable data sources. With multiple pre-trained models available as starting points, developers can also readily customize them for their domain-specific use cases, such as Information Technology, Human Resource help assistants, and Research \u0026 Development research assistants.\n\n### Terms of use\n\nThe use of this model is governed by\nthe [NVIDIA NeMo Foundational Models Evaluation License Agreement](https://registry.ngc.nvidia.com/orgs/ohlfw0olaadg/teams/ea-participants/resources/nemo_foundational_models_evaluation_license/files)\n\n### References(s)\n\nThe NVIDIA Retrieval QA Embedding model is meant to be deployed using the [NeMo Retriever Embedding Microservice](https://registry.ngc.nvidia.com/orgs/ohlfw0olaadg/teams/ea-participants/containers/nemo-retriever-embedding-microservice). Check out the microservice documentation for more details.\n\n[Technical Blog](https://developer.nvidia.com/blog/build-enterprise-retrieval-augmented-generation-apps-with-nvidia-retrieval-qa-embedding-model/)\n\n### Intended use\n\nThe NVIDIA Retrieval QA Embedding model is most suitable for users who want to build a question and answer application over a large text corpus, leveraging the latest dense retrieval technologies.\n\n### Model Architecture\n\n**Architecture Type:** Transformer \u003cbr\u003e\n**Network Architecture:** Fine-tuned E5-Large-Unsupervised retriever \u003cbr\u003e\n**Embedding Dimension:** 1024 \u003cbr\u003e\n**Parameter Count:** 335 million \u003cbr\u003e\n\nThe NVIDIA Retrieval QA Embedding Model is a transformer encoder - a finetuned version of [E5-Large-Unsupervised](https://huggingface.co/intfloat/e5-large-unsupervised), with 24 layers and an embedding size of 1024, which is trained on private and public datasets as described in the Dataset and Training section. It supports a maximum input of 512 tokens.\n\nEmbedding models for text retrieval are typically trained using a bi-encoder architecture. This involves encoding a pair of sentences (for example, query and chunked passages) independently using the embedding model. Contrastive learning is used to maximize the similarity between the query and the passage that contains the answer, while minimizing the similarity between the query and sampled negative passages not useful to answer the question.\n\n### Input\n\n**Input Type:** text \u003cbr\u003e\n**Input Format:** list of strings \u003cbr\u003e\n\n### Output\n\n**Output Type:** floats \u003cbr\u003e\n**Output Format:** list of float arrays, each array containing the embeddings for the corresponding input string. \u003cbr\u003e\n\n### Model Version(s)\n\nNVIDIA Retrieval QA Embedding Model-4.0\n\n## Training Dataset \u0026 Evaluation\n\n### Training Dataset\n\nThe development of large-scale public open-QA datasets has enabled tremendous progress in powerful embedding models. However, one popular dataset named [MSMARCO](https://microsoft.github.io/msmarco/) restricts ‌commercial licensing, limiting the use of these models in commercial settings. To address this, we created our own internal open-domain QA dataset to train a commercially-viable embedding model. For NVIDIA proprietary data collection, we searched the passages from web logs and selected a collection of passages relevant to customer use cases for annotation by the NVIDIA internal data annotation team.\n\nTo minimize the redundancy in our data collection process, we selected samples that maximized relevancy distance scores and increased diversity in the data. The pretrained [E5-Large-Unsupervised](https://huggingface.co/intfloat/e5-large-unsupervised) embedding model was fine-tuned with contrastive learning with the prefix of “query:” for questions and “passage:” for context passages. Specifically, a mixture of English language datasets are used including our proprietary dataset, along with selected samples from commercially-viable public datasets. The AdamW optimizer is employed, incorporating 300 warm-up steps and 1e-6 learning rate with linear annealing schedule.\n\nThe training dataset details are as follows:\n\n**Use Case**: Information retrieval for question and answering over text documents. \u003cbr\u003e\n\n**Data Sources**: \u003cbr\u003e\n- Public datasets licensed for commercial use. \u003cbr\u003e\n- Text from public websites. \u003cbr\u003e\n- Annotations created by NVIDIA’s internal team.\u003cbr\u003e\n\n**Language**: English (US) \u003cbr\u003e\n\n**Domains**: Knowledge, Description, Numeric (unit, time), Entity, Location, Person \u003cbr\u003e\n\n**Volume**: 40k internal proprietary samples, 200k samples from public dataset \u003cbr\u003e\n\n**High Level Schema**: \u003cbr\u003e\n- query: question text \u003cbr\u003e\n- doc: full document that contains the answer \u003cbr\u003e\n- chunk: section of the document that contains the answer \u003cbr\u003e\n- relevancy label: rating of how relevant the passage is to the question \u003cbr\u003e\n- span: exact token range in the chunk that contains the answer \u003cbr\u003e\n\n### Evaluation Results\n\nWe evaluated the NVIDIA Retrieval QA Embedding Model in comparison to literature open \u0026 commercial retriever models on academic benchmarks - [NQ](https://huggingface.co/datasets/BeIR/nq), [HotpotQA](https://huggingface.co/datasets/hotpot_qa) and [FiQA(Finance Q\u0026A)](https://huggingface.co/datasets/BeIR/fiqa) from BeIR benchmark, and the [TechQA(Tech Support Q\u0026A)](https://arxiv.org/pdf/1911.02984v1.pdf) dataset. In this benchmark, the metric used was [Recall@5](https://en.wikipedia.org/wiki/Precision_and_recall).\n\n| Open \u0026 Commercial Retrieval Models | Average Recall@5 on NQ, HotpotQA, FiQA, TechQA dataset|\n|-----------------------------|----------------------------|\n| NVIDIA Retrieval QA | 57.37% |\n| E5-Large_unsupervised | 45.58% |\n| BM25 | 39.97% |\n\nWe also evaluated our embedding model with real internal customer datasets from telco, IT, consulting, and energy industries. The metric was Recall@5, to emulate a retrieval augmented generation (RAG) scenario where we would provide the top five most relevant passages as context in the prompt for the LLM model that is going to respond to the question. We compared our model’s information retrieval accuracy to a number of well-known embedding models made available by the AI community, including ones trained on non-commercial dataset (which are marked with \"*\").\n\n| Retrieval Model | Average Recall@5 on Internal Customer Datasets |\n|-----------------------------|-----------------------------|\n| NVIDIA Retrieval QA | 74.4% |\n| DRAGON* | 72.7% |\n| E5-Large* | 71.7% |\n| BGE* | 71.1% |\n| GTR* | 71.0% |\n| Contriever* | 69.0% |\n| GTE* | 63.9% |\n| E5-Large_unsupervised | 61.6% |\n| BM25 | 55.6% |\n\n## Ethical Considerations\n\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety \u0026 Security, and Privacy Subcards [here](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/ai-foundation/models/nvolve-29k/bias). Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).\n\n### Special Training Data Considerations\n\nThe model was trained on the data that may contain toxic language and societal biases originally crawled from the Internet. Therefore, the model may amplify those biases, for example, associating certain genders with certain social stereotypes."])</script><script>self.__next_f.push([1,"24f:T641,| Field | Response |\n|:-------------------|:---------|\n|Intended Application \u0026 Domain: | Passage and query embedding for question and answer retrieval |\n|Model Type: | Transformer encoder |\n|Intended User: | Generative AI creators working with conversational AI models. |\n|Output: | Text embedding (An array of float numbers, providing a dense vector representation for the input text) |\n|Describe how the model works: | The transformer encoder transforms the tokenized input text into a dense vector representation. |\n|Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of: | Not Applicable |\nVerified to have met prescribed NVIDIA quality standards: | Yes\n|Performance Metrics: | Accuracy, Throughput, and Latency |\n|Potential Known Risks: | The model was trained on the data that may contain toxic language and societal biases originally crawled from the Internet. Therefore, the model may amplify those biases, for example, associating certain genders with certain social stereotypes. |\n|Licensing: | [NVIDIA NeMo Foundational Models Evaluation License Agreement](https://registry.ngc.nvidia.com/orgs/ohlfw0olaadg/teams/ea-participants/resources/nemo_foundational_models_evaluation_license/files)|\n|Technical Limitations: | The model's maximum context length is 512 tokens. Texts longer than maximum length must either be chunked or truncated.|250:T5af,| Field | Response |\n|:-------------------|:---------|\n|Generatable or reverse engineerable personally-identifiable information (PII)? | None |\n|Was consent obtained for any PII used? | Not Applicable |\n|IPII used to create this model? | None |\n|How often is dataset reviewed? "])</script><script>self.__next_f.push([1," | Before Release |\n|Is a mechanism in place to honor data subject right of access or deletion of personal data? | No |\n|If PII collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable |\n|If PII collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable |\n|If PII collected for the development of this AI model, was it minimized to only what was required? | Not Applicable |\n|Is there provenance for all datasets used in training? | Yes |\n|Are we able to identify and trace source of dataset? | Yes |\n|Does data labeling (annotation, metadata) comply with privacy laws? | Yes | \n|Is data compliant with data subject requests for data correction or removal, if such a request was made? | No, not possible with externally-sourced data|253:{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"}\n254:{\"name\":\"CC-BY-NC-4.0\",\"url\":\"https://spdx.org/licenses/CC-BY-NC-4.0\"}\n252:{\"title\":\"NVIDIA NIM API for nvidia/embed-qa-4\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nvidia-embed-qa-4 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":\"$253\",\"license\":\"$254\"}\n256:{\"url\":\"https://integrate.api.nvidia.com/v1\"}\n255:[\"$256\"]\n25a:[\"Embeddings\"]\n262:{\"type\":\"string\"}\n264:{\"type\":\"string\"}\n263:{\"type\":\"array\",\"items\":\"$264\"}\n261:[\"$262\",\"$263\"]\n260:{\"description\":\"Input text to embed. Max length depends on model.\",\"oneOf\":\"$261\",\"minLength\":1,\"maxLength\":4096,\"title\":\"Input\"}\n265:{\"type\":\"string\",\"description\":\"ID of the embedding model.\",\"example\":\"nvidia/embed-qa-4\",\"title\":\"Model\"}\n267:[\"passage\",\"query\"]\n266:{\"type\":\"string\",\"enum\":\"$267\",\"description\":\"nvidia/embed-qa-4 and E5 models operate in `passage` or `query` mode, and thus require the "])</script><script>self.__next_f.push([1,"`input_type` parameter. `passage` is used when generating embeddings during indexing. `query` is used when generating embeddings during querying. It is very important to use the correct `input_type`. Failure to do so will result in large drops in retrieval accuracy. As an alternative, it is possible to add the `-query` or `-passage` suffix to the `model` parameter like `nvidia/embed-qa-4-query` and not use the `input_type` field at all for OpenAI API compliance. Please note that the GTE model _does not_ accept the `input_type` parameter since both the query and passage are processed in the same way.\",\"title\":\"Input Type\"}\n269:[\"float\",\"base64\"]\n268:{\"type\":\"string\",\"description\":\"The format to return the embeddings in.\",\"enum\":\"$269\",\"default\":\"float\",\"title\":\"Encoding Format\"}\n26b:[\"NONE\",\"START\",\"END\"]\n26a:{\"type\":\"string\",\"description\":\"Specifies how inputs longer than the maximum token length of the model are handled. Passing `START` discards the start of the input. `END` discards the end of the input. In both cases, input is discarded until the remaining input is exactly the maximum input token length for the model. If `NONE` is selected, when the input exceeds the maximum input token length an error will be returned.\",\"enum\":\"$26b\",\"default\":\"NONE\",\"title\":\"Truncate\"}\n26c:{\"type\":\"string\",\"description\":\"Not implemented, but provided for API compliance. This field is ignored.\",\"title\":\"User\"}\n25f:{\"input\":\"$260\",\"model\":\"$265\",\"input_type\":\"$266\",\"encoding_format\":\"$268\",\"truncate\":\"$26a\",\"user\":\"$26c\"}\n26d:[\"input\",\"model\"]\n25e:{\"type\":\"object\",\"properties\":\"$25f\",\"required\":\"$26d\"}\n25d:{\"schema\":\"$25e\"}\n25c:{\"application/json\":\"$25d\"}\n25b:{\"required\":true,\"content\":\"$25c\"}\n271:T57ca,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [ -0.0175628662109375, -0.035247802734375, 0.044586181640625, -0.034881591796875, 0.0277099609375, -0.01152801513671875, 0.00598907470703125, -0.026031494140625, 0.023895263671875, -0.00873565673828125, 0.01508331298828125, 0.052337646484375, -0.00351715087890625, -0.005550384521484375, 0.018768310546875, -0.0264739990234375, -0.0262298583984375, -0.01971435546875, 0.040008544921875, -0.019195556640625, -0.01557159423828125, -0.023468017578125, 0.038665771484375, 0.0242919921875, -0.03631591796875, 0.041290283203125, 0.01129150390625, 0.0255584716796875, 0.007411956787109375, 0.005908966064453125, 0.03338623046875, -0.0628662109375, 0.01464080810546875, 0.046295166015625, 0.0084381103515625, -0.040802001953125, -0.039215087890625, -0.0233612060546875, -0.0482177734375, 0.01471710205078125, -0.0226287841796875, -0.030731201171875, 0.020599365234375, 0.09307861328125, 0.00038886070251464844, -0.00815582275390625, 0.00786590576171875, -0.06182861328125, 0.0242919921875, -0.05206298828125, 0.0254058837890625, 0.0254669189453125, 0.06396484375, 0.0178375244140625, -0.032135009765625, 0.0184326171875, 0.0609130859375, 0.071044921875, -0.0197906494140625, -0.036956787109375, 0.0290985107421875, -0.078857421875, -0.0015840530395507812, -0.0198974609375, -0.033355712890625, 0.0084686279296875, 0.002529144287109375, -0.044403076171875, 0.015869140625, -0.0236358642578125, -0.0005974769592285156, -0.020599365234375, 0.017730712890625, -0.02294921875, 0.043548583984375, -0.050689697265625, -0.035186767578125, 0.02178955078125, 0.013580322265625, 0.0615234375, -0.01495361328125, 0.00872802734375, 0.0018701553344726562, -0.0248260498046875, 0.00992584228515625, -0.01361083984375, 0.0709228515625, 0.05712890625, -0.015869140625, -0.0007290840148925781, -0.033355712890625, -0.0078582763671875, 0.052398681640625, -0.00958251953125, -0.0095062255859375, 0.01354217529296875, -0.0176239013671875, 0.0283966064453125, 0.00710296630859375, -0.0223541259765625, -0.01190185546875, -0.01412200927734375, 0.00897979736328125, 0.037567138671875, 0.0009965896606445312, 0.0201873779296875, 0.03729248046875, 0.0362548828125, -0.03704833984375, 0.02081298828125, -0.0182342529296875, -0.00568389892578125, 0.00821685791015625, 0.01947021484375, 0.00022363662719726562, -0.005825042724609375, -0.020233154296875, -0.0277099609375, -0.013671875, 0.051666259765625, 0.018157958984375, 0.01837158203125, 0.0301055908203125, 0.01177215576171875, -0.04156494140625, 0.06207275390625, -0.016143798828125, -0.031646728515625, -0.061248779296875, -0.0059051513671875, -0.0248260498046875, -0.033599853515625, 0.05169677734375, -0.0244903564453125, 0.0013380050659179688, -0.0263671875, -0.048248291015625, -0.0509033203125, 0.0272369384765625, -0.0011968612670898438, 0.0088348388671875, -0.0264892578125, -0.003879547119140625, 0.00629425048828125, 0.062103271484375, 0.0308380126953125, 0.01055908203125, -0.00960540771484375, 0.00162506103515625, 0.01461029052734375, -0.03765869140625, 0.0278167724609375, -0.0286407470703125, 0.0019464492797851562, 0.0177764892578125, -0.0230865478515625, -0.01122283935546875, 0.060272216796875, -0.027984619140625, -0.0152435302734375, 0.010498046875, 0.0106201171875, 0.0055084228515625, 0.018341064453125, -0.004058837890625, -0.0212554931640625, 0.04119873046875, -0.0213165283203125, 0.05401611328125, -0.006816864013671875, -0.047576904296875, -0.00685882568359375, 0.0654296875, 0.007343292236328125, 0.040435791015625, -0.006717681884765625, 0.0321044921875, -0.013458251953125, -0.026214599609375, 0.051971435546875, -0.035888671875, -0.0175628662109375, 0.043731689453125, 0.0458984375, 0.03125, 0.0158843994140625, -0.03900146484375, 0.01239013671875, -0.005107879638671875, 0.0009927749633789062, -0.063232421875, -0.0579833984375, 0.057342529296875, -0.0162506103515625, 0.0318603515625, 0.019989013671875, 0.0159759521484375, 0.00249481201171875, 0.019012451171875, -0.055450439453125, -0.038726806640625, 0.01552581787109375, 0.0128326416015625, -0.0222930908203125, 0.005161285400390625, 0.0555419921875, 0.010986328125, -0.017059326171875, -0.018829345703125, 0.004001617431640625, -0.041351318359375, 0.00566864013671875, -0.02593994140625, -0.0026683807373046875, -0.064697265625, -0.0137176513671875, 0.00902557373046875, 0.0211181640625, -0.0041656494140625, 0.039886474609375, -0.01611328125, 0.05657958984375, 0.0609130859375, 0.03179931640625, 0.0052947998046875, -0.005023956298828125, -0.061309814453125, -0.015960693359375, 0.002635955810546875, 0.0257720947265625, -0.055450439453125, -0.00621795654296875, 0.047271728515625, 0.000043451786041259766, 0.042022705078125, -0.023681640625, 0.047088623046875, -0.03814697265625, 0.0726318359375, 0.03350830078125, -0.02716064453125, 0.0011434555053710938, 0.0233612060546875, 0.04119873046875, 0.01219940185546875, 0.00318145751953125, 0.0142364501953125, 0.02935791015625, 0.05499267578125, 0.0142822265625, -0.012359619140625, 0.054046630859375, 0.0023365020751953125, -0.06536865234375, -0.005855560302734375, -0.012115478515625, 0.027984619140625, -0.048675537109375, -0.01800537109375, 0.08160400390625, -0.0292816162109375, 0.03192138671875, 0.0171966552734375, 0.019073486328125, 0.0187225341796875, -0.036376953125, 0.02142333984375, -0.048065185546875, 0.062286376953125, -0.051544189453125, 0.00608062744140625, -0.00591278076171875, -0.01114654541015625, -0.0303802490234375, 0.00946044921875, 0.031951904296875, 0.0052032470703125, -0.0245513916015625, -0.03326416015625, -0.02191162109375, 0.03497314453125, -0.0350341796875, 0.0035991668701171875, -0.022186279296875, -0.04864501953125, -0.026214599609375, -0.00824737548828125, 0.00009715557098388672, 0.01511383056640625, -0.02301025390625, 0.00632476806640625, 0.040863037109375, -0.034759521484375, -0.0321044921875, 0.0027599334716796875, 0.00803375244140625, 0.01885986328125, -0.0233001708984375, 0.0233154296875, 0.010955810546875, 0.0154571533203125, -0.057769775390625, 0.0240020751953125, 0.03411865234375, -0.03131103515625, -0.0181732177734375, 0.0028209686279296875, 0.01348114013671875, -0.06884765625, 0.01412200927734375, 0.0288848876953125, -0.02581787109375, 0.042816162109375, 0.0017719268798828125, -0.02093505859375, 0.0259552001953125, -0.023040771484375, 0.011474609375, -0.01812744140625, -0.01099395751953125, 0.03326416015625, 0.01380157470703125, -0.0147705078125, -0.01702880859375, -0.0026912689208984375, 0.01593017578125, -0.0187225341796875, 0.027008056640625, 0.0180206298828125, 0.0164642333984375, -0.036041259765625, -0.01776123046875, -0.01363372802734375, 0.050201416015625, -0.0019388198852539062, -0.030487060546875, 0.0010833740234375, 0.007232666015625, -0.02899169921875, 0.054901123046875, -0.004779815673828125, -0.01276397705078125, 0.035186767578125, -0.023681640625, 0.0029201507568359375, 0.0106201171875, -0.0272674560546875, 0.06707763671875, 0.004215240478515625, 0.04339599609375, -0.01486968994140625, -0.037445068359375, 0.0182952880859375, 0.0102386474609375, -0.01468658447265625, 0.005229949951171875, 0.0017652511596679688, -0.02203369140625, 0.0117340087890625, 0.036895751953125, -0.01132965087890625, -0.0045318603515625, -0.00010925531387329102, -0.031982421875, 0.0010833740234375, 0.02093505859375, -0.02874755859375, -0.05914306640625, -0.055206298828125, 0.075439453125, 0.0081024169921875, 0.0020694732666015625, -0.011077880859375, 0.04119873046875, -0.071044921875, -0.0150146484375, -0.031982421875, 0.009185791015625, 0.029541015625, 0.01136016845703125, 0.04522705078125, 0.01318359375, 0.01824951171875, -0.00812530517578125, -0.03973388671875, -0.0474853515625, -0.0242462158203125, 0.0013980865478515625, 0.01291656494140625, 0.00904083251953125, -0.005100250244140625, 0.01007843017578125, 0.00856781005859375, -0.021484375, -0.011688232421875, -0.045379638671875, 0.0084381103515625, 0.00750732421875, 0.01369476318359375, -0.0264739990234375, 0.0273895263671875, 0.00806427001953125, 0.02557373046875, 0.016571044921875, -0.005931854248046875, -0.028045654296875, 0.0035648345947265625, 0.0477294921875, 0.0216522216796875, -0.052215576171875, 0.0433349609375, 0.0225677490234375, 0.00005179643630981445, -0.0291595458984375, -0.021453857421875, -0.0025234222412109375, 0.006702423095703125, 0.00835418701171875, 0.00324249267578125, 0.039093017578125, 0.0169677734375, -0.02435302734375, -0.000054836273193359375, 0.032623291015625, -0.0325927734375, -0.01000213623046875, 0.051849365234375, 0.039306640625, -0.0281829833984375, 0.03369140625, 0.05340576171875, 0.0165863037109375, -0.00799560546875, -0.08905029296875, -0.02764892578125, -0.0322265625, -0.0279693603515625, -0.0007419586181640625, -0.021484375, -0.004199981689453125, -0.0205535888671875, -0.019500732421875, -0.01027679443359375, -0.024658203125, -0.01139068603515625, 0.0157318115234375, 0.04345703125, -0.01532745361328125, 0.028472900390625, -0.038909912109375, 0.005680084228515625, 0.0246429443359375, -0.01023101806640625, -0.0345458984375, -0.066650390625, 0.006328582763671875, 0.014984130859375, -0.002597808837890625, 0.042694091796875, 0.0108489990234375, -0.0198211669921875, 0.0120849609375, -0.008544921875, -0.045318603515625, 0.005558013916015625, -0.01230621337890625, 0.0253448486328125, -0.044586181640625, -0.0198211669921875, -0.0151519775390625, 0.019439697265625, -0.00640106201171875, 0.0232086181640625, -0.04296875, -0.0191497802734375, 0.002742767333984375, 0.045654296875, -0.0179595947265625, 0.07421875, 0.0421142578125, 0.0227813720703125, -0.01195526123046875, 0.0162506103515625, -0.0121917724609375, -0.0574951171875, 0.051910400390625, -0.0009026527404785156, 0.024658203125, 0.01375579833984375, -0.01454925537109375, 0.0163421630859375, 0.028076171875, 0.0301055908203125, 0.0103607177734375, -0.03955078125, -0.003047943115234375, 0.047943115234375, 0.0169830322265625, 0.03558349609375, 0.036102294921875, 0.037078857421875, 0.023895263671875, -0.01000213623046875, 0.0193939208984375, 0.03814697265625, -0.0648193359375, -0.0247039794921875, -0.0149078369140625, -0.007965087890625, -0.044952392578125, 0.0248260498046875, 0.00893402099609375, 0.0302734375, -0.002170562744140625, -0.0293121337890625, 0.01580810546875, 0.03387451171875, 0.032257080078125, -0.037689208984375, -0.0086517333984375, -0.0103912353515625, -0.0379638671875, 0.0018854141235351562, -0.027191162109375, -0.07049560546875, 0.0009918212890625, 0.031494140625, 0.01355743408203125, 0.01190185546875, 0.006916046142578125, 0.024566650390625, -0.0012769699096679688, -0.01203155517578125, 0.032440185546875, -0.034210205078125, 0.03302001953125, -0.02581787109375, -0.041900634765625, 0.00798797607421875, 0.0013275146484375, -0.0472412109375, 0.033843994140625, -0.04864501953125, 0.01702880859375, -0.0016794204711914062, 0.0253448486328125, -0.0789794921875, -0.055450439453125, 0.00783538818359375, 0.004093170166015625, 0.0233001708984375, -0.052001953125, 0.0201416015625, -0.052337646484375, 0.005352020263671875, 0.031646728515625, 0.0277252197265625, 0.04071044921875, 0.01430511474609375, 0.009490966796875, -0.0421142578125, 0.07415771484375, 0.0382080078125, -0.019744873046875, 0.004650115966796875, 0.0178680419921875, -0.0268707275390625, -0.0109710693359375, 0.004146575927734375, -0.00307464599609375, 0.0546875, -0.028411865234375, -0.03424072265625, 0.038177490234375, -0.0372314453125, 0.03271484375, 0.003421783447265625, -0.007236480712890625, 0.0210113525390625, -0.02947998046875, 0.017730712890625, -0.05706787109375, 0.034027099609375, 0.006542205810546875, -0.007518768310546875, -0.006488800048828125, 0.034332275390625, 0.024566650390625, 0.0264739990234375, 0.0074920654296875, -0.006229400634765625, 0.06500244140625, -0.0206451416015625, -0.009063720703125, -0.012481689453125, -0.044281005859375, 0.01044464111328125, -0.0023441314697265625, 0.045379638671875, 0.003055572509765625, -0.0290374755859375, -0.0233154296875, -0.0345458984375, -0.05059814453125, 0.04388427734375, 0.0146026611328125, 0.0113067626953125, -0.01285552978515625, -0.00010091066360473633, -0.02874755859375, 0.013275146484375, 0.000316619873046875, 0.0192413330078125, 0.0279083251953125, -0.024993896484375, 0.0362548828125, -0.01552581787109375, -0.0137176513671875, 0.044158935546875, 0.06707763671875, -0.0111541748046875, 0.01326751708984375, -0.004917144775390625, -0.0078582763671875, 0.057647705078125, -0.0130615234375, -0.030487060546875, 0.00970458984375, 0.0168304443359375, 0.032501220703125, -0.033050537109375, -0.049530029296875, 0.020172119140625, -0.0255584716796875, 0.03271484375, -0.01384735107421875, -0.00930023193359375, -0.028961181640625, -0.0292816162109375, 0.02642822265625, -0.01337432861328125, -0.003292083740234375, 0.045684814453125, -0.00431060791015625, 0.0499267578125, 0.036407470703125, -0.00847625732421875, -0.051910400390625, -0.051239013671875, 0.033447265625, -0.0272979736328125, 0.03948974609375, 0.0672607421875, -0.0183258056640625, 0.0433349609375, 0.034881591796875, 0.0338134765625, -0.036102294921875, 0.01422882080078125, 0.02996826171875, 0.079345703125, -0.00345611572265625, 0.020263671875, -0.0188751220703125, 0.0024471282958984375, 0.12042236328125, 0.052886962890625, 0.0256805419921875, -0.044464111328125, -0.05120849609375, -0.0279388427734375, 0.0195770263671875, -0.005786895751953125, -0.01042938232421875, 0.00942230224609375, -0.0123748779296875, 0.061676025390625, -0.00473785400390625, 0.01107025146484375, 0.0299072265625, -0.0079803466796875, -0.003692626953125, 0.08154296875, 0.0176849365234375, -0.00885009765625, 0.019683837890625, -0.01186370849609375, -0.06512451171875, -0.017822265625, 0.0162811279296875, -0.0219879150390625, 0.023529052734375, -0.0206451416015625, -0.01546478271484375, 0.023529052734375, -0.003368377685546875, 0.016876220703125, -0.004364013671875, 0.0194091796875, 0.021697998046875, -0.0034313201904296875, 0.0009307861328125, 0.062744140625, -0.037353515625, 0.004169464111328125, 0.0147857666015625, 0.00099945068359375, -0.0340576171875, 0.0216522216796875, -0.0211029052734375, 0.00015842914581298828, 0.01358795166015625, -0.00841522216796875, -0.033203125, -0.01336669921875, 0.06884765625, -0.0025482177734375, -0.0120697021484375, -0.030975341796875, 0.0457763671875, -0.0087738037109375, 0.0220947265625, -0.017822265625, 0.0025806427001953125, -0.017852783203125, 0.0217742919921875, -0.059906005859375, 0.02294921875, 0.0033359527587890625, -0.035125732421875, 0.01311492919921875, -0.0094757080078125, 0.0226287841796875, -0.03997802734375, -0.05792236328125, 0.0246734619140625, -0.01450347900390625, -0.019256591796875, -0.04718017578125, 0.031890869140625, -0.018157958984375, -0.00597381591796875, 0.004894256591796875, -0.05950927734375, -0.0204010009765625, -0.038360595703125, 0.001361846923828125, 0.0255584716796875, -0.01404571533203125, -0.0295257568359375, -0.0223388671875, 0.06024169921875, -0.004787445068359375, -0.047607421875, 0.0157012939453125, -0.032806396484375, 0.00218963623046875, 0.0152587890625, -0.040435791015625, -0.017791748046875, 0.050384521484375, -0.0592041015625, 0.0002696514129638672, 0.01259613037109375, 0.04437255859375, 0.0333251953125, 0.02099609375, 0.00826263427734375, 0.0236358642578125, 0.0389404296875, -0.0009889602661132812, -0.01366424560546875, -0.03240966796875, -0.007049560546875, 0.019500732421875, 0.0249176025390625, -0.00579833984375, -0.0262603759765625, -0.07281494140625, -0.051666259765625, 0.037628173828125, -0.01885986328125, -0.0394287109375, 0.00856781005859375, 0.024993896484375, 0.008880615234375, 0.019287109375, 0.06622314453125, 0.023223876953125, 0.043731689453125, 0.06640625, 0.012298583984375, 0.0595703125, -0.02691650390625, -0.02642822265625, 0.036102294921875, 0.037445068359375, 0.0033168792724609375, -0.01294708251953125, 0.00689697265625, 0.005443572998046875, 0.01410675048828125, 0.045867919921875, -0.050628662109375, -0.014984130859375, 0.0390625, -0.070556640625, -0.0242156982421875, 0.007061004638671875, -0.0152587890625, -0.00603485107421875, 0.0004017353057861328, -0.0261688232421875, 0.032928466796875, 0.050567626953125, -0.0177001953125, -0.0259552001953125, 0.0263214111328125, -0.040130615234375, 0.031982421875, 0.028717041015625, 0.0164337158203125, -0.01131439208984375, 0.0271759033203125, 0.00821685791015625, 0.024383544921875, -0.052734375, 0.06640625, -0.0063018798828125, -0.016143798828125, -0.0223541259765625, 0.0189056396484375, -0.042266845703125, -0.00015866756439208984, 0.007030487060546875, 0.0294342041015625, 0.0091705322265625, -0.030914306640625, 0.0313720703125, -0.05584716796875, 0.01409149169921875, -0.0035228729248046875, -0.0231475830078125, 0.023101806640625, 0.01361083984375, 0.02545166015625, 0.052825927734375, -0.0521240234375, 0.0298004150390625, 0.004116058349609375, -0.041259765625, -0.04693603515625, -0.002010345458984375, 0.0311431884765625, -0.0270538330078125, 0.03143310546875, 0.0022449493408203125, 0.0207977294921875, -0.0002307891845703125, -0.037017822265625, -0.0272216796875, 0.0009908676147460938, 0.033538818359375, -0.053253173828125, -0.01450347900390625, -0.00408935546875, 0.00792694091796875, -0.003879547119140625, 0.02825927734375, 0.0291748046875, -0.01529693603515625, 0.01099395751953125, 0.0889892578125, 0.03289794921875, 0.0460205078125, 0.0306549072265625, -0.031982421875, -0.036651611328125, -0.04254150390625, -0.0194091796875, 0.03759765625, 0.00347900390625, 0.06036376953125, -0.025421142578125, -0.00423431396484375, -0.003814697265625, 0.02587890625, -0.041900634765625, -0.0789794921875, -0.01227569580078125, -0.0214996337890625, 0.037445068359375, -0.02197265625, -0.019256591796875, -0.01383209228515625, -0.026947021484375, -0.00579833984375, -0.03240966796875, -0.050201416015625, 0.006320953369140625, -0.02374267578125, -0.026702880859375, -0.004070281982421875, -0.01287078857421875, -0.0229034423828125, -0.04022216796875, -0.035888671875, -0.020172119140625, 0.00264739990234375, 0.005374908447265625, -0.029815673828125, -0.033905029296875, 0.059844970703125, 0.0062713623046875, 0.041900634765625, 0.04034423828125, 0.00513458251953125, 0.0091705322265625, -0.04864501953125, -0.00991058349609375, -7.152557373046875e-7, 0.000392913818359375, -0.00359344482421875, 0.00519561767578125, -0.016815185546875, -0.00007575750350952148, 0.03521728515625, -0.0215911865234375, -0.023162841796875, -0.0240325927734375, -0.018951416015625, 0.030975341796875, -0.01117706298828125, -0.008331298828125, -0.01529693603515625, -0.05206298828125, 0.0012445449829101562, 0.0023136138916015625, 0.064208984375, -0.007354736328125, -0.03753662109375, 0.0007100105285644531, 0.0036830902099609375, 0.00608062744140625, -0.01541900634765625, -0.021087646484375, 0.0225067138671875, 0.00518035888671875, 0.05523681640625, 0.00641632080078125, -0.0830078125, 0.061248779296875, 0.045318603515625, -0.064208984375, -0.023193359375, 0.0083770751953125, -0.004787445068359375, 0.04266357421875, 0.0689697265625, 0.0213470458984375, -0.06317138671875, 0.028350830078125, 0.0179290771484375, -0.03997802734375, -0.00140380859375, 0.031646728515625, -0.019287109375, -0.0117950439453125, 0.0109100341796875, 0.01340484619140625, -0.046295166015625, -0.03973388671875, 0.01273345947265625, 0.036468505859375, -0.01299285888671875, -0.00962066650390625, 0.032135009765625, 0.0130157470703125, 0.00604248046875, 0.0191192626953125, -0.0135650634765625, -0.003337860107421875, 0.0054779052734375, -0.058868408203125, -0.00634002685546875, -0.0305938720703125, -0.019989013671875, 0.024444580078125, 0.0097503662109375, 0.07818603515625, -0.03997802734375, 0.0227813720703125, -0.0131683349609375, -0.007534027099609375, -0.045257568359375, 0.0631103515625, 0.02471923828125, 0.038665771484375, 0.01050567626953125, 0.0038280487060546875, 0.046875, -0.033203125, 0.035308837890625, -0.037872314453125, -0.0115203857421875, -0.039520263671875, -0.0279693603515625, -0.01386260986328125, -0.0177001953125, 0.039520263671875, -0.016387939453125, -0.03021240234375, 0.0186920166015625, -0.008392333984375, 0.01145172119140625, -0.0269775390625, -0.016082763671875, 0.0242919921875, -0.039794921875, 0.0809326171875, -0.01479339599609375, -0.039520263671875, -0.00608062744140625, 0.030303955078125],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/embed-qa-4\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"270:{\"name\":\"Embedding vector for text input\",\"requestJson\":\"{\\n \\\"input\\\": \\\"What is the capital of France?\\\",\\n \\\"model\\\": \\\"nvidia/embed-qa-4\\\",\\n \\\"input_type\\\": \\\"query\\\",\\n \\\"encoding_format\\\": \\\"float\\\",\\n \\\"truncate\\\": \\\"NONE\\\"\\n}\\n\",\"responseJson\":\"$271\"}\n26f:[\"$270\"]\n274:{\"curl\":\"curl -X POST https://integrate.api.nvidia.com/v1/embeddings \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"input\\\": [\\\"\u003c%- request.input.replaceAll('\\\"', '\\\\\\\\\\\"').replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\\\"],\\n \\\"model\\\": \\\"nvidia/embed-qa-4\\\",\\n \\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\",\\n \\\"encoding_format\\\": \\\"\u003c%- request.encoding_format %\u003e\\\",\\n \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"\\n }'\\n\",\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n api_key=\\\"$NVIDIA_API_KEY\\\",\\n base_url=\\\"https://integrate.api.nvidia.com/v1\\\"\\n)\\n\\nresponse = client.embeddings.create(\\n input=[\u003c%- JSON.stringify(request.input) %\u003e],\\n model=\\\"nvidia/embed-qa-4\\\",\\n encoding_format=\\\"\u003c%- request.encoding_format %\u003e\\\",\\n extra_body={\\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\", \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"}\\n)\\n\\nprint(response.data[0].embedding)\\n\",\"langchain\":\"from langchain_nvidia_ai_endpoints import NVIDIAEmbeddings\\n\\nclient = NVIDIAEmbeddings(\\n model=\\\"NV-Embed-QA\\\", \\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n truncate=\\\"\u003c%- request.truncate %\u003e\\\", \\n )\\n\\nembedding = client.embed_query(\u003c%- JSON.stringify(request.input) %\u003e)\\nprint(embedding)\\n\"}\n275:T57ca,"])</script><script>self.__next_f.push([1,"{\n \"object\": \"list\",\n \"data\": [\n {\n \"object\": \"embedding\",\n \"embedding\": [ -0.0175628662109375, -0.035247802734375, 0.044586181640625, -0.034881591796875, 0.0277099609375, -0.01152801513671875, 0.00598907470703125, -0.026031494140625, 0.023895263671875, -0.00873565673828125, 0.01508331298828125, 0.052337646484375, -0.00351715087890625, -0.005550384521484375, 0.018768310546875, -0.0264739990234375, -0.0262298583984375, -0.01971435546875, 0.040008544921875, -0.019195556640625, -0.01557159423828125, -0.023468017578125, 0.038665771484375, 0.0242919921875, -0.03631591796875, 0.041290283203125, 0.01129150390625, 0.0255584716796875, 0.007411956787109375, 0.005908966064453125, 0.03338623046875, -0.0628662109375, 0.01464080810546875, 0.046295166015625, 0.0084381103515625, -0.040802001953125, -0.039215087890625, -0.0233612060546875, -0.0482177734375, 0.01471710205078125, -0.0226287841796875, -0.030731201171875, 0.020599365234375, 0.09307861328125, 0.00038886070251464844, -0.00815582275390625, 0.00786590576171875, -0.06182861328125, 0.0242919921875, -0.05206298828125, 0.0254058837890625, 0.0254669189453125, 0.06396484375, 0.0178375244140625, -0.032135009765625, 0.0184326171875, 0.0609130859375, 0.071044921875, -0.0197906494140625, -0.036956787109375, 0.0290985107421875, -0.078857421875, -0.0015840530395507812, -0.0198974609375, -0.033355712890625, 0.0084686279296875, 0.002529144287109375, -0.044403076171875, 0.015869140625, -0.0236358642578125, -0.0005974769592285156, -0.020599365234375, 0.017730712890625, -0.02294921875, 0.043548583984375, -0.050689697265625, -0.035186767578125, 0.02178955078125, 0.013580322265625, 0.0615234375, -0.01495361328125, 0.00872802734375, 0.0018701553344726562, -0.0248260498046875, 0.00992584228515625, -0.01361083984375, 0.0709228515625, 0.05712890625, -0.015869140625, -0.0007290840148925781, -0.033355712890625, -0.0078582763671875, 0.052398681640625, -0.00958251953125, -0.0095062255859375, 0.01354217529296875, -0.0176239013671875, 0.0283966064453125, 0.00710296630859375, -0.0223541259765625, -0.01190185546875, -0.01412200927734375, 0.00897979736328125, 0.037567138671875, 0.0009965896606445312, 0.0201873779296875, 0.03729248046875, 0.0362548828125, -0.03704833984375, 0.02081298828125, -0.0182342529296875, -0.00568389892578125, 0.00821685791015625, 0.01947021484375, 0.00022363662719726562, -0.005825042724609375, -0.020233154296875, -0.0277099609375, -0.013671875, 0.051666259765625, 0.018157958984375, 0.01837158203125, 0.0301055908203125, 0.01177215576171875, -0.04156494140625, 0.06207275390625, -0.016143798828125, -0.031646728515625, -0.061248779296875, -0.0059051513671875, -0.0248260498046875, -0.033599853515625, 0.05169677734375, -0.0244903564453125, 0.0013380050659179688, -0.0263671875, -0.048248291015625, -0.0509033203125, 0.0272369384765625, -0.0011968612670898438, 0.0088348388671875, -0.0264892578125, -0.003879547119140625, 0.00629425048828125, 0.062103271484375, 0.0308380126953125, 0.01055908203125, -0.00960540771484375, 0.00162506103515625, 0.01461029052734375, -0.03765869140625, 0.0278167724609375, -0.0286407470703125, 0.0019464492797851562, 0.0177764892578125, -0.0230865478515625, -0.01122283935546875, 0.060272216796875, -0.027984619140625, -0.0152435302734375, 0.010498046875, 0.0106201171875, 0.0055084228515625, 0.018341064453125, -0.004058837890625, -0.0212554931640625, 0.04119873046875, -0.0213165283203125, 0.05401611328125, -0.006816864013671875, -0.047576904296875, -0.00685882568359375, 0.0654296875, 0.007343292236328125, 0.040435791015625, -0.006717681884765625, 0.0321044921875, -0.013458251953125, -0.026214599609375, 0.051971435546875, -0.035888671875, -0.0175628662109375, 0.043731689453125, 0.0458984375, 0.03125, 0.0158843994140625, -0.03900146484375, 0.01239013671875, -0.005107879638671875, 0.0009927749633789062, -0.063232421875, -0.0579833984375, 0.057342529296875, -0.0162506103515625, 0.0318603515625, 0.019989013671875, 0.0159759521484375, 0.00249481201171875, 0.019012451171875, -0.055450439453125, -0.038726806640625, 0.01552581787109375, 0.0128326416015625, -0.0222930908203125, 0.005161285400390625, 0.0555419921875, 0.010986328125, -0.017059326171875, -0.018829345703125, 0.004001617431640625, -0.041351318359375, 0.00566864013671875, -0.02593994140625, -0.0026683807373046875, -0.064697265625, -0.0137176513671875, 0.00902557373046875, 0.0211181640625, -0.0041656494140625, 0.039886474609375, -0.01611328125, 0.05657958984375, 0.0609130859375, 0.03179931640625, 0.0052947998046875, -0.005023956298828125, -0.061309814453125, -0.015960693359375, 0.002635955810546875, 0.0257720947265625, -0.055450439453125, -0.00621795654296875, 0.047271728515625, 0.000043451786041259766, 0.042022705078125, -0.023681640625, 0.047088623046875, -0.03814697265625, 0.0726318359375, 0.03350830078125, -0.02716064453125, 0.0011434555053710938, 0.0233612060546875, 0.04119873046875, 0.01219940185546875, 0.00318145751953125, 0.0142364501953125, 0.02935791015625, 0.05499267578125, 0.0142822265625, -0.012359619140625, 0.054046630859375, 0.0023365020751953125, -0.06536865234375, -0.005855560302734375, -0.012115478515625, 0.027984619140625, -0.048675537109375, -0.01800537109375, 0.08160400390625, -0.0292816162109375, 0.03192138671875, 0.0171966552734375, 0.019073486328125, 0.0187225341796875, -0.036376953125, 0.02142333984375, -0.048065185546875, 0.062286376953125, -0.051544189453125, 0.00608062744140625, -0.00591278076171875, -0.01114654541015625, -0.0303802490234375, 0.00946044921875, 0.031951904296875, 0.0052032470703125, -0.0245513916015625, -0.03326416015625, -0.02191162109375, 0.03497314453125, -0.0350341796875, 0.0035991668701171875, -0.022186279296875, -0.04864501953125, -0.026214599609375, -0.00824737548828125, 0.00009715557098388672, 0.01511383056640625, -0.02301025390625, 0.00632476806640625, 0.040863037109375, -0.034759521484375, -0.0321044921875, 0.0027599334716796875, 0.00803375244140625, 0.01885986328125, -0.0233001708984375, 0.0233154296875, 0.010955810546875, 0.0154571533203125, -0.057769775390625, 0.0240020751953125, 0.03411865234375, -0.03131103515625, -0.0181732177734375, 0.0028209686279296875, 0.01348114013671875, -0.06884765625, 0.01412200927734375, 0.0288848876953125, -0.02581787109375, 0.042816162109375, 0.0017719268798828125, -0.02093505859375, 0.0259552001953125, -0.023040771484375, 0.011474609375, -0.01812744140625, -0.01099395751953125, 0.03326416015625, 0.01380157470703125, -0.0147705078125, -0.01702880859375, -0.0026912689208984375, 0.01593017578125, -0.0187225341796875, 0.027008056640625, 0.0180206298828125, 0.0164642333984375, -0.036041259765625, -0.01776123046875, -0.01363372802734375, 0.050201416015625, -0.0019388198852539062, -0.030487060546875, 0.0010833740234375, 0.007232666015625, -0.02899169921875, 0.054901123046875, -0.004779815673828125, -0.01276397705078125, 0.035186767578125, -0.023681640625, 0.0029201507568359375, 0.0106201171875, -0.0272674560546875, 0.06707763671875, 0.004215240478515625, 0.04339599609375, -0.01486968994140625, -0.037445068359375, 0.0182952880859375, 0.0102386474609375, -0.01468658447265625, 0.005229949951171875, 0.0017652511596679688, -0.02203369140625, 0.0117340087890625, 0.036895751953125, -0.01132965087890625, -0.0045318603515625, -0.00010925531387329102, -0.031982421875, 0.0010833740234375, 0.02093505859375, -0.02874755859375, -0.05914306640625, -0.055206298828125, 0.075439453125, 0.0081024169921875, 0.0020694732666015625, -0.011077880859375, 0.04119873046875, -0.071044921875, -0.0150146484375, -0.031982421875, 0.009185791015625, 0.029541015625, 0.01136016845703125, 0.04522705078125, 0.01318359375, 0.01824951171875, -0.00812530517578125, -0.03973388671875, -0.0474853515625, -0.0242462158203125, 0.0013980865478515625, 0.01291656494140625, 0.00904083251953125, -0.005100250244140625, 0.01007843017578125, 0.00856781005859375, -0.021484375, -0.011688232421875, -0.045379638671875, 0.0084381103515625, 0.00750732421875, 0.01369476318359375, -0.0264739990234375, 0.0273895263671875, 0.00806427001953125, 0.02557373046875, 0.016571044921875, -0.005931854248046875, -0.028045654296875, 0.0035648345947265625, 0.0477294921875, 0.0216522216796875, -0.052215576171875, 0.0433349609375, 0.0225677490234375, 0.00005179643630981445, -0.0291595458984375, -0.021453857421875, -0.0025234222412109375, 0.006702423095703125, 0.00835418701171875, 0.00324249267578125, 0.039093017578125, 0.0169677734375, -0.02435302734375, -0.000054836273193359375, 0.032623291015625, -0.0325927734375, -0.01000213623046875, 0.051849365234375, 0.039306640625, -0.0281829833984375, 0.03369140625, 0.05340576171875, 0.0165863037109375, -0.00799560546875, -0.08905029296875, -0.02764892578125, -0.0322265625, -0.0279693603515625, -0.0007419586181640625, -0.021484375, -0.004199981689453125, -0.0205535888671875, -0.019500732421875, -0.01027679443359375, -0.024658203125, -0.01139068603515625, 0.0157318115234375, 0.04345703125, -0.01532745361328125, 0.028472900390625, -0.038909912109375, 0.005680084228515625, 0.0246429443359375, -0.01023101806640625, -0.0345458984375, -0.066650390625, 0.006328582763671875, 0.014984130859375, -0.002597808837890625, 0.042694091796875, 0.0108489990234375, -0.0198211669921875, 0.0120849609375, -0.008544921875, -0.045318603515625, 0.005558013916015625, -0.01230621337890625, 0.0253448486328125, -0.044586181640625, -0.0198211669921875, -0.0151519775390625, 0.019439697265625, -0.00640106201171875, 0.0232086181640625, -0.04296875, -0.0191497802734375, 0.002742767333984375, 0.045654296875, -0.0179595947265625, 0.07421875, 0.0421142578125, 0.0227813720703125, -0.01195526123046875, 0.0162506103515625, -0.0121917724609375, -0.0574951171875, 0.051910400390625, -0.0009026527404785156, 0.024658203125, 0.01375579833984375, -0.01454925537109375, 0.0163421630859375, 0.028076171875, 0.0301055908203125, 0.0103607177734375, -0.03955078125, -0.003047943115234375, 0.047943115234375, 0.0169830322265625, 0.03558349609375, 0.036102294921875, 0.037078857421875, 0.023895263671875, -0.01000213623046875, 0.0193939208984375, 0.03814697265625, -0.0648193359375, -0.0247039794921875, -0.0149078369140625, -0.007965087890625, -0.044952392578125, 0.0248260498046875, 0.00893402099609375, 0.0302734375, -0.002170562744140625, -0.0293121337890625, 0.01580810546875, 0.03387451171875, 0.032257080078125, -0.037689208984375, -0.0086517333984375, -0.0103912353515625, -0.0379638671875, 0.0018854141235351562, -0.027191162109375, -0.07049560546875, 0.0009918212890625, 0.031494140625, 0.01355743408203125, 0.01190185546875, 0.006916046142578125, 0.024566650390625, -0.0012769699096679688, -0.01203155517578125, 0.032440185546875, -0.034210205078125, 0.03302001953125, -0.02581787109375, -0.041900634765625, 0.00798797607421875, 0.0013275146484375, -0.0472412109375, 0.033843994140625, -0.04864501953125, 0.01702880859375, -0.0016794204711914062, 0.0253448486328125, -0.0789794921875, -0.055450439453125, 0.00783538818359375, 0.004093170166015625, 0.0233001708984375, -0.052001953125, 0.0201416015625, -0.052337646484375, 0.005352020263671875, 0.031646728515625, 0.0277252197265625, 0.04071044921875, 0.01430511474609375, 0.009490966796875, -0.0421142578125, 0.07415771484375, 0.0382080078125, -0.019744873046875, 0.004650115966796875, 0.0178680419921875, -0.0268707275390625, -0.0109710693359375, 0.004146575927734375, -0.00307464599609375, 0.0546875, -0.028411865234375, -0.03424072265625, 0.038177490234375, -0.0372314453125, 0.03271484375, 0.003421783447265625, -0.007236480712890625, 0.0210113525390625, -0.02947998046875, 0.017730712890625, -0.05706787109375, 0.034027099609375, 0.006542205810546875, -0.007518768310546875, -0.006488800048828125, 0.034332275390625, 0.024566650390625, 0.0264739990234375, 0.0074920654296875, -0.006229400634765625, 0.06500244140625, -0.0206451416015625, -0.009063720703125, -0.012481689453125, -0.044281005859375, 0.01044464111328125, -0.0023441314697265625, 0.045379638671875, 0.003055572509765625, -0.0290374755859375, -0.0233154296875, -0.0345458984375, -0.05059814453125, 0.04388427734375, 0.0146026611328125, 0.0113067626953125, -0.01285552978515625, -0.00010091066360473633, -0.02874755859375, 0.013275146484375, 0.000316619873046875, 0.0192413330078125, 0.0279083251953125, -0.024993896484375, 0.0362548828125, -0.01552581787109375, -0.0137176513671875, 0.044158935546875, 0.06707763671875, -0.0111541748046875, 0.01326751708984375, -0.004917144775390625, -0.0078582763671875, 0.057647705078125, -0.0130615234375, -0.030487060546875, 0.00970458984375, 0.0168304443359375, 0.032501220703125, -0.033050537109375, -0.049530029296875, 0.020172119140625, -0.0255584716796875, 0.03271484375, -0.01384735107421875, -0.00930023193359375, -0.028961181640625, -0.0292816162109375, 0.02642822265625, -0.01337432861328125, -0.003292083740234375, 0.045684814453125, -0.00431060791015625, 0.0499267578125, 0.036407470703125, -0.00847625732421875, -0.051910400390625, -0.051239013671875, 0.033447265625, -0.0272979736328125, 0.03948974609375, 0.0672607421875, -0.0183258056640625, 0.0433349609375, 0.034881591796875, 0.0338134765625, -0.036102294921875, 0.01422882080078125, 0.02996826171875, 0.079345703125, -0.00345611572265625, 0.020263671875, -0.0188751220703125, 0.0024471282958984375, 0.12042236328125, 0.052886962890625, 0.0256805419921875, -0.044464111328125, -0.05120849609375, -0.0279388427734375, 0.0195770263671875, -0.005786895751953125, -0.01042938232421875, 0.00942230224609375, -0.0123748779296875, 0.061676025390625, -0.00473785400390625, 0.01107025146484375, 0.0299072265625, -0.0079803466796875, -0.003692626953125, 0.08154296875, 0.0176849365234375, -0.00885009765625, 0.019683837890625, -0.01186370849609375, -0.06512451171875, -0.017822265625, 0.0162811279296875, -0.0219879150390625, 0.023529052734375, -0.0206451416015625, -0.01546478271484375, 0.023529052734375, -0.003368377685546875, 0.016876220703125, -0.004364013671875, 0.0194091796875, 0.021697998046875, -0.0034313201904296875, 0.0009307861328125, 0.062744140625, -0.037353515625, 0.004169464111328125, 0.0147857666015625, 0.00099945068359375, -0.0340576171875, 0.0216522216796875, -0.0211029052734375, 0.00015842914581298828, 0.01358795166015625, -0.00841522216796875, -0.033203125, -0.01336669921875, 0.06884765625, -0.0025482177734375, -0.0120697021484375, -0.030975341796875, 0.0457763671875, -0.0087738037109375, 0.0220947265625, -0.017822265625, 0.0025806427001953125, -0.017852783203125, 0.0217742919921875, -0.059906005859375, 0.02294921875, 0.0033359527587890625, -0.035125732421875, 0.01311492919921875, -0.0094757080078125, 0.0226287841796875, -0.03997802734375, -0.05792236328125, 0.0246734619140625, -0.01450347900390625, -0.019256591796875, -0.04718017578125, 0.031890869140625, -0.018157958984375, -0.00597381591796875, 0.004894256591796875, -0.05950927734375, -0.0204010009765625, -0.038360595703125, 0.001361846923828125, 0.0255584716796875, -0.01404571533203125, -0.0295257568359375, -0.0223388671875, 0.06024169921875, -0.004787445068359375, -0.047607421875, 0.0157012939453125, -0.032806396484375, 0.00218963623046875, 0.0152587890625, -0.040435791015625, -0.017791748046875, 0.050384521484375, -0.0592041015625, 0.0002696514129638672, 0.01259613037109375, 0.04437255859375, 0.0333251953125, 0.02099609375, 0.00826263427734375, 0.0236358642578125, 0.0389404296875, -0.0009889602661132812, -0.01366424560546875, -0.03240966796875, -0.007049560546875, 0.019500732421875, 0.0249176025390625, -0.00579833984375, -0.0262603759765625, -0.07281494140625, -0.051666259765625, 0.037628173828125, -0.01885986328125, -0.0394287109375, 0.00856781005859375, 0.024993896484375, 0.008880615234375, 0.019287109375, 0.06622314453125, 0.023223876953125, 0.043731689453125, 0.06640625, 0.012298583984375, 0.0595703125, -0.02691650390625, -0.02642822265625, 0.036102294921875, 0.037445068359375, 0.0033168792724609375, -0.01294708251953125, 0.00689697265625, 0.005443572998046875, 0.01410675048828125, 0.045867919921875, -0.050628662109375, -0.014984130859375, 0.0390625, -0.070556640625, -0.0242156982421875, 0.007061004638671875, -0.0152587890625, -0.00603485107421875, 0.0004017353057861328, -0.0261688232421875, 0.032928466796875, 0.050567626953125, -0.0177001953125, -0.0259552001953125, 0.0263214111328125, -0.040130615234375, 0.031982421875, 0.028717041015625, 0.0164337158203125, -0.01131439208984375, 0.0271759033203125, 0.00821685791015625, 0.024383544921875, -0.052734375, 0.06640625, -0.0063018798828125, -0.016143798828125, -0.0223541259765625, 0.0189056396484375, -0.042266845703125, -0.00015866756439208984, 0.007030487060546875, 0.0294342041015625, 0.0091705322265625, -0.030914306640625, 0.0313720703125, -0.05584716796875, 0.01409149169921875, -0.0035228729248046875, -0.0231475830078125, 0.023101806640625, 0.01361083984375, 0.02545166015625, 0.052825927734375, -0.0521240234375, 0.0298004150390625, 0.004116058349609375, -0.041259765625, -0.04693603515625, -0.002010345458984375, 0.0311431884765625, -0.0270538330078125, 0.03143310546875, 0.0022449493408203125, 0.0207977294921875, -0.0002307891845703125, -0.037017822265625, -0.0272216796875, 0.0009908676147460938, 0.033538818359375, -0.053253173828125, -0.01450347900390625, -0.00408935546875, 0.00792694091796875, -0.003879547119140625, 0.02825927734375, 0.0291748046875, -0.01529693603515625, 0.01099395751953125, 0.0889892578125, 0.03289794921875, 0.0460205078125, 0.0306549072265625, -0.031982421875, -0.036651611328125, -0.04254150390625, -0.0194091796875, 0.03759765625, 0.00347900390625, 0.06036376953125, -0.025421142578125, -0.00423431396484375, -0.003814697265625, 0.02587890625, -0.041900634765625, -0.0789794921875, -0.01227569580078125, -0.0214996337890625, 0.037445068359375, -0.02197265625, -0.019256591796875, -0.01383209228515625, -0.026947021484375, -0.00579833984375, -0.03240966796875, -0.050201416015625, 0.006320953369140625, -0.02374267578125, -0.026702880859375, -0.004070281982421875, -0.01287078857421875, -0.0229034423828125, -0.04022216796875, -0.035888671875, -0.020172119140625, 0.00264739990234375, 0.005374908447265625, -0.029815673828125, -0.033905029296875, 0.059844970703125, 0.0062713623046875, 0.041900634765625, 0.04034423828125, 0.00513458251953125, 0.0091705322265625, -0.04864501953125, -0.00991058349609375, -7.152557373046875e-7, 0.000392913818359375, -0.00359344482421875, 0.00519561767578125, -0.016815185546875, -0.00007575750350952148, 0.03521728515625, -0.0215911865234375, -0.023162841796875, -0.0240325927734375, -0.018951416015625, 0.030975341796875, -0.01117706298828125, -0.008331298828125, -0.01529693603515625, -0.05206298828125, 0.0012445449829101562, 0.0023136138916015625, 0.064208984375, -0.007354736328125, -0.03753662109375, 0.0007100105285644531, 0.0036830902099609375, 0.00608062744140625, -0.01541900634765625, -0.021087646484375, 0.0225067138671875, 0.00518035888671875, 0.05523681640625, 0.00641632080078125, -0.0830078125, 0.061248779296875, 0.045318603515625, -0.064208984375, -0.023193359375, 0.0083770751953125, -0.004787445068359375, 0.04266357421875, 0.0689697265625, 0.0213470458984375, -0.06317138671875, 0.028350830078125, 0.0179290771484375, -0.03997802734375, -0.00140380859375, 0.031646728515625, -0.019287109375, -0.0117950439453125, 0.0109100341796875, 0.01340484619140625, -0.046295166015625, -0.03973388671875, 0.01273345947265625, 0.036468505859375, -0.01299285888671875, -0.00962066650390625, 0.032135009765625, 0.0130157470703125, 0.00604248046875, 0.0191192626953125, -0.0135650634765625, -0.003337860107421875, 0.0054779052734375, -0.058868408203125, -0.00634002685546875, -0.0305938720703125, -0.019989013671875, 0.024444580078125, 0.0097503662109375, 0.07818603515625, -0.03997802734375, 0.0227813720703125, -0.0131683349609375, -0.007534027099609375, -0.045257568359375, 0.0631103515625, 0.02471923828125, 0.038665771484375, 0.01050567626953125, 0.0038280487060546875, 0.046875, -0.033203125, 0.035308837890625, -0.037872314453125, -0.0115203857421875, -0.039520263671875, -0.0279693603515625, -0.01386260986328125, -0.0177001953125, 0.039520263671875, -0.016387939453125, -0.03021240234375, 0.0186920166015625, -0.008392333984375, 0.01145172119140625, -0.0269775390625, -0.016082763671875, 0.0242919921875, -0.039794921875, 0.0809326171875, -0.01479339599609375, -0.039520263671875, -0.00608062744140625, 0.030303955078125],\n \"index\": 0\n }\n ],\n \"model\": \"nvidia/embed-qa-4\",\n \"usage\": {\n \"prompt_tokens\": 0,\n \"total_tokens\": 0\n }\n}\n"])</script><script>self.__next_f.push([1,"273:{\"title\":\"Synchronous requests\",\"requestEjs\":\"$274\",\"response\":\"$275\"}\n272:[\"$273\"]\n26e:{\"name\":\"Create Text Embedding\",\"description\":\"Generates an embedding vector from the provided text\\nusing a specified model. The embedding can be returned\\nin either float array or base64-encoded format.\\n\",\"path\":\"create\",\"examples\":\"$26f\",\"templates\":\"$272\"}\n27c:{\"type\":\"string\",\"example\":\"list\"}\n27d:{\"type\":\"array\",\"items\":\"$244\"}\n27e:{\"type\":\"string\",\"example\":\"nre-001\"}\n281:{\"type\":\"integer\",\"example\":0}\n282:{\"type\":\"integer\",\"example\":0}\n280:{\"prompt_tokens\":\"$281\",\"total_tokens\":\"$282\"}\n27f:{\"type\":\"object\",\"description\":\"Not implemented, but provided for API compliance. This field will contain zeros.\",\"properties\":\"$280\"}\n27b:{\"object\":\"$27c\",\"data\":\"$27d\",\"model\":\"$27e\",\"usage\":\"$27f\"}\n27a:{\"type\":\"object\",\"properties\":\"$27b\"}\n279:{\"schema\":\"$27a\"}\n278:{\"application/json\":\"$279\"}\n277:{\"description\":\"Successful response\",\"content\":\"$278\"}\n288:{\"type\":\"string\"}\n289:{\"type\":\"string\"}\n28a:{\"type\":\"object\",\"additionalProperties\":true}\n28b:{\"type\":\"string\"}\n287:{\"object\":\"$288\",\"message\":\"$289\",\"detail\":\"$28a\",\"type\":\"$28b\"}\n286:{\"type\":\"object\",\"properties\":\"$287\"}\n285:{\"schema\":\"$286\"}\n284:{\"application/json\":\"$285\"}\n283:{\"description\":\"Bad request\",\"content\":\"$284\"}\n276:{\"200\":\"$277\",\"400\":\"$283\"}\n259:{\"tags\":\"$25a\",\"summary\":\"Creates an embedding vector from the input text.\",\"operationId\":\"create_embedding\",\"requestBody\":\"$25b\",\"x-nvai-meta\":\"$26e\",\"responses\":\"$276\"}\n258:{\"post\":\"$259\"}\n257:{\"/embeddings\":\"$258\"}\n28e:{\"type\":\"http\",\"scheme\":\"bearer\"}\n28d:{\"Token\":\"$28e\"}\n28f:{\"EmbeddingObject\":\"$244\"}\n28c:{\"securitySchemes\":\"$28d\",\"schemas\":\"$28f\"}\n251:{\"openapi\":\"3.1.0\",\"info\":\"$252\",\"servers\":\"$255\",\"paths\":\"$257\",\"components\":\"$28c\"}\n293:[\"system\",\"context\",\"user\",\"assistant\"]\n292:{\"description\":\"The role of the message author.\",\"enum\":\"$293\",\"title\":\"Role\",\"type\":\"string\"}\n294:{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"type\":\"string\"}\n291:{\"role\":\"$292\",\"content\":\"$294\"}\n295:[\"role\",\"cont"])</script><script>self.__next_f.push([1,"ent\"]\n290:{\"additionalProperties\":false,\"properties\":\"$291\",\"required\":\"$295\",\"title\":\"Message\",\"type\":\"object\"}\n298:{\"type\":\"string\",\"description\":\"Error type\"}\n299:{\"type\":\"string\",\"description\":\"Error title\"}\n29a:{\"type\":\"integer\",\"description\":\"Error status code\"}\n29b:{\"type\":\"string\",\"description\":\"Detailed information about the error\"}\n29c:{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"}\n29d:{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}\n297:{\"type\":\"$298\",\"title\":\"$299\",\"status\":\"$29a\",\"detail\":\"$29b\",\"instance\":\"$29c\",\"requestId\":\"$29d\"}\n29e:[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"]\n296:{\"properties\":\"$297\",\"type\":\"object\",\"required\":\"$29e\",\"title\":\"InvokeError\"}\n2a1:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\n2a5:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\n2a7:[\"$290\"]\n2a9:{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}\n2a8:[\"$2a9\"]\n2a6:{\"allOf\":\"$2a7\",\"description\":\"A chat completion message generated by the model.\",\"examples\":\"$2a8\"}\n2ad:[\"stop\",\"length\"]\n2ac:{\"enum\":\"$2ad\",\"type\":\"string\"}\n2ae:{\"type\":\"null\"}\n2ab:[\"$2ac\",\"$2ae\"]\n2af:[\"stop\"]\n2aa:{\"anyOf\":\"$2ab\",\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":\"$2af\",\"title\":\"Finish Reason\"}\n2a4:{\"index\":\"$2a5\",\"message\":\"$2a6\",\"finish_reason\":\"$2aa\"}\n2b0:[\"index\",\"message\"]\n2a3:{\"properties\":\"$2a4\",\"required\":\"$2b0\",\"title\":\"Choice\",\"type\":\"object\"}\n2a2:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$2a3\",\"title\":\"Choices\",\"type\":\"array\"}\n2b6:[25]\n2b5:{\"description\":\"Number of tokens in the generated completion.\",\"ex"])</script><script>self.__next_f.push([1,"amples\":\"$2b6\",\"title\":\"Completion Tokens\",\"type\":\"integer\"}\n2b8:[9]\n2b7:{\"description\":\"Number of tokens in the prompt.\",\"examples\":\"$2b8\",\"title\":\"Prompt Tokens\",\"type\":\"integer\"}\n2ba:[34]\n2b9:{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":\"$2ba\",\"title\":\"Total Tokens\",\"type\":\"integer\"}\n2b4:{\"completion_tokens\":\"$2b5\",\"prompt_tokens\":\"$2b7\",\"total_tokens\":\"$2b9\"}\n2bb:[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"]\n2b3:{\"properties\":\"$2b4\",\"required\":\"$2bb\",\"title\":\"Usage\",\"type\":\"object\"}\n2b2:[\"$2b3\"]\n2b1:{\"allOf\":\"$2b2\",\"description\":\"Usage statistics for the completion request.\"}\n2a0:{\"id\":\"$2a1\",\"choices\":\"$2a2\",\"usage\":\"$2b1\"}\n2bc:[\"id\",\"choices\",\"usage\"]\n29f:{\"properties\":\"$2a0\",\"required\":\"$2bc\",\"title\":\"ChatCompletion\",\"type\":\"object\"}\n2bf:{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"}\n2c3:{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"}\n2c5:[\"$290\"]\n2c7:{\"content\":\"Ah,\",\"role\":\"assistant\"}\n2c6:[\"$2c7\"]\n2c4:{\"allOf\":\"$2c5\",\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":\"$2c6\"}\n2cb:[\"stop\",\"length\"]\n2ca:{\"enum\":\"$2cb\",\"type\":\"string\"}\n2cc:{\"type\":\"null\"}\n2c9:[\"$2ca\",\"$2cc\"]\n2c8:{\"anyOf\":\"$2c9\",\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}\n2c2:{\"index\":\"$2c3\",\"delta\":\"$2c4\",\"finish_reason\":\"$2c8\"}\n2cd:[\"index\",\"delta\"]\n2c1:{\"properties\":\"$2c2\",\"required\":\"$2cd\",\"title\":\"ChoiceChunk\",\"type\":\"object\"}\n2c0:{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":\"$2c1\",\"title\":\"Choices\",\"type\":\"array\"}\n2be:{\"id\":\"$2bf\",\"choices\":\"$2c0\"}\n2ce:[\"id\",\"choices\"]\n2bd:{\"properties\":\"$2be\","])</script><script>self.__next_f.push([1,"\"required\":\"$2ce\",\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}\n2d1:{\"type\":\"string\",\"title\":\"Model\",\"default\":\"qwen/qwen2.5-coder-32b-instruct\"}\n2d5:{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}\n2d4:[\"$2d5\"]\n2d3:[\"$2d4\"]\n2d2:{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":\"$2d3\",\"items\":\"$290\",\"title\":\"Messages\",\"type\":\"array\"}\n2d6:{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Temperature\",\"type\":\"number\"}\n2d7:{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"}\n2d8:{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4000,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"}\n2db:{\"maximum\":18446744073709552000,\"minimum\":0,\"type\":\"integer\"}\n2dc:{\"type\":\"null\"}\n2da:[\"$2db\",\"$2dc\"]\n2dd:[42]\n2d9:{\"anyOf\":\"$2da\",\"default\":null,\"description\":\"If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.\",\"exa"])</script><script>self.__next_f.push([1,"mples\":\"$2dd\",\"title\":\"Seed\"}\n2de:{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"}\n2e2:{\"type\":\"string\"}\n2e1:{\"items\":\"$2e2\",\"type\":\"array\"}\n2e3:{\"type\":\"string\"}\n2e4:{\"type\":\"null\"}\n2e0:[\"$2e1\",\"$2e3\",\"$2e4\"]\n2df:{\"anyOf\":\"$2e0\",\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}\n2d0:{\"model\":\"$2d1\",\"messages\":\"$2d2\",\"temperature\":\"$2d6\",\"top_p\":\"$2d7\",\"max_tokens\":\"$2d8\",\"seed\":\"$2d9\",\"stream\":\"$2de\",\"stop\":\"$2df\"}\n2e5:[\"messages\"]\n2cf:{\"additionalProperties\":false,\"properties\":\"$2d0\",\"required\":\"$2e5\",\"title\":\"ChatRequest\",\"type\":\"object\"}\n2e6:Tf69,"])</script><script>self.__next_f.push([1,"# Model Overview\n\n## Description:\nQwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). As of now, Qwen2.5-Coder has covered six mainstream model sizes, 0.5, 1.5, 3, 7, 14, 32 billion parameters, to meet the needs of different developers. Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:\n* Significant improvements in code generation, code reasoning and code fixing. Increased training tokens to 5.5 trillion including source code, text-code grounding, Synthetic data, etc. Qwen2.5-Coder-32B has become the current state-of-the-art open-source codeLLM, with its coding abilities matching those of GPT-4o.\n* A more comprehensive foundation for real-world applications such as Code Agents. Not only enhancing coding capabilities but also maintaining its strengths in mathematics and general competencies.\n* Long-context support up to 32K tokens.\n\nThis model is ready for commercial/non-commercial use.\n\n## Third-Party Community Consideration\nThis model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA [Qwen2.5-Coder-32B-Instruct Model Card](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct).\n\n## License/Terms of Use\nQwen/Qwen2.5-Coder-32B-Instruct is licensed under the [Apache 2.0 License](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct/blob/main/LICENSE)\n\n## References:\n[Blog](https://qwenlm.github.io/blog/qwen2.5-coder-family/), [Github](https://github.com/QwenLM/Qwen2.5), [Technical Report](https://arxiv.org/abs/2409.12186)\n\n## Model Architecture:\n**Architecture Type:** Transformer \u003cbr\u003e\n**Network Architecture:** Qwen2.5-Coder-32B-Instruct\n\n## Input:\n**Input Type(s):** Text \u003cbr\u003e\n**Input Format(s):** String \u003cbr\u003e\n**Input Parameters:** 1D\n\n## Output:\n**Output Type(s):** Text \u003cbr\u003e\n**Output Format:** String \u003cbr\u003e\n**Output Parameters:** 1D\n\n## Model Version(s):\nQwen2.5-Coder-32B-Instruct\n\n## Training, Testing, and Evaluation Datasets:\n\n## Training Dataset:\n**Link:** Unknown \u003cbr\u003e\n**Data Collection Method by dataset:** Hybrid: Automated, Human \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Automated, Synthetic \u003cbr\u003e\n**Properties:** The training dataset contains over 5.5 trillion tokens total across 92 programming languages with a mixture ratio of 70% Code, 20% Text, 10% Math, sourced from GitHub repositories, Pull Requests, Commits, Jupyter Notebooks, and Kaggle datasets.\n\n## Testing Dataset:\n**Link:** Unknown \u003cbr\u003e\n**Data Collection Method by dataset:** Unknown \u003cbr\u003e\n**Labeling Method by dataset:** Unknown \u003cbr\u003e\n**Properties:** Unknown\n\n## Evaluation Dataset:\n**Link:** See evaluation section of the [Hugging Face Qwen2.5-Coder-32B-Instruct Model Card](https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct#evaluation--performance) \u003cbr\u003e\n**Data Collection Method by dataset:** Hybrid: Human, Automated \u003cbr\u003e\n**Labeling Method by dataset:** Hybrid: Automated, Human \u003cbr\u003e\n**Properties:** The evaluation datasets consist of multiple benchmarks including HumanEval with 164 Python programming tasks, MBPP with 974 programming problems, LiveCodeBench with over 600 coding problems, and additional benchmarks covering code generation, completion, reasoning and debugging capabilities.\n\n## Inference:\n**Engine:** TensorRT-LLM \u003cbr\u003e\n**Test Hardware:** NVIDIA L40S\n\n## Ethical Considerations:\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. \n\nPlease report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/)."])</script><script>self.__next_f.push([1,"2e9:{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"}\n2ea:{\"name\":\"Apache 2.0\",\"url\":\"https://apache.org/licenses/LICENSE-2.0\"}\n2e8:{\"title\":\"NVIDIA NIM API for qwen/qwen2.5-coder-32b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/qwen-qwen2_5-coder-32b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":\"$2e9\",\"license\":\"$2ea\"}\n2ec:{\"url\":\"https://integrate.api.nvidia.com/v1/\"}\n2eb:[\"$2ec\"]\n2f0:[\"Chat\"]\n2f3:{\"schema\":\"$2cf\"}\n2f2:{\"application/json\":\"$2f3\"}\n2f1:{\"content\":\"$2f2\",\"required\":true}\n2f7:{\"schema\":\"$29f\"}\n2f8:{\"schema\":\"$2bd\"}\n2f6:{\"application/json\":\"$2f7\",\"text/event-stream\":\"$2f8\"}\n2f5:{\"description\":\"Invocation is fulfilled\",\"content\":\"$2f6\"}\n2fc:{}\n2fd:{}\n2fb:{\"example\":\"$2fc\",\"schema\":\"$2fd\"}\n2fa:{\"application/json\":\"$2fb\"}\n300:{\"type\":\"string\",\"format\":\"uuid\"}\n2ff:{\"description\":\"requestId required for pooling\",\"schema\":\"$300\"}\n302:{\"type\":\"string\"}\n301:{\"description\":\"Invocation status\",\"schema\":\"$302\"}\n2fe:{\"NVCF-REQID\":\"$2ff\",\"NVCF-STATUS\":\"$301\"}\n2f9:{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":\"$2fa\",\"headers\":\"$2fe\"}\n306:{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n305:{\"schema\":\"$296\",\"example\":\"$306\"}\n304:{\"application/json\":\"$305\"}\n303:{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":\"$304\"}\n30a:{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}\n309:{\"schema\":\"$296\",\"example\":\"$30a\"}\n308:{\"app"])</script><script>self.__next_f.push([1,"lication/json\":\"$309\"}\n307:{\"description\":\"The invocation ended with an error.\",\"content\":\"$308\"}\n2f4:{\"200\":\"$2f5\",\"202\":\"$2f9\",\"422\":\"$303\",\"500\":\"$307\"}\n30d:{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n30e:{\"name\":\"What can I see at NVIDIA's GPU Technology Conference?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What can I see at NVIDIA's GPU Technology Conference?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n30c:[\"$30d\",\"$30e\"]\n"])</script><script>self.__next_f.push([1,"311:{\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n base_url = \\\"https://integrate.api.nvidia.com/v1\\\",\\n api_key = \\\"$NVIDIA_API_KEY\\\"\\n)\\n\\ncompletion = client.chat.completions.create(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in completion:\\n if chunk.choices[0].delta.content is not None:\\n print(chunk.choices[0].delta.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nprint(completion.choices[0].message)\\n\u003c% } %\u003e\\n\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"import OpenAI from 'openai';\\n\\nconst openai = new OpenAI({\\n apiKey: '$NVIDIA_API_KEY',\\n baseURL: 'https://integrate.api.nvidia.com/v1',\\n})\\n\\nasync function main() {\\n const completion = await openai.chat.completions.create({\\n model: \\\"\u003c%- request.model %\u003e\\\",\\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature: \u003c%- request.temperature %\u003e,\\n top_p: \u003c%- request.top_p %\u003e,\\n max_tokens: \u003c%- request.max_tokens %\u003e,\\n stream: \u003c%- request.stream %\u003e,\\n })\\n \u003c% if (request.stream) { %\u003e\\n for await (const chunk of completion) {\\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\\n }\\n \u003c% } else { %\u003e\\n process.stdout.write(completion.choices[0]?.message?.content);\\n \u003c% } %\u003e\\n}\\n\\nmain();\",\"curl\":\"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature %\u003e, \\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e \\n }'\\n\"}\n"])</script><script>self.__next_f.push([1,"310:{\"title\":\"No Streaming\",\"requestEjs\":\"$311\",\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}\n30f:[\"$310\"]\n30b:{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":\"$30c\",\"templates\":\"$30f\"}\n2ef:{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":\"$2f0\",\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":\"$2f1\",\"responses\":\"$2f4\",\"x-nvai-meta\":\"$30b\"}\n2ee:{\"post\":\"$2ef\"}\n2ed:{\"/chat/completions\":\"$2ee\"}\n314:[]\n313:{\"Token\":\"$314\"}\n312:[\"$313\"]\n317:{\"type\":\"http\",\"scheme\":\"bearer\"}\n316:{\"Token\":\"$317\"}\n318:{\"Errors\":\"$296\",\"ChatCompletion\":\"$29f\",\"ChatCompletionChunk\":\"$2bd\",\"ChatRequest\":\"$2cf\",\"Choice\":\"$2a3\",\"ChoiceChunk\":\"$2c1\",\"Message\":\"$290\",\"Usage\":\"$2b3\"}\n315:{\"securitySchemes\":\"$316\",\"schemas\":\"$318\"}\n2e7:{\"openapi\":\"3.1.0\",\"info\":\"$2e8\",\"servers\":\"$2eb\",\"paths\":\"$2ed\",\"security\":\"$312\",\"components\":\"$315\"}\n"])</script><script>self.__next_f.push([1,"19:[[\"$\",\"$L21\",null,{\"flag\":\"killswitches/quick-actions.enabled\",\"initialData\":true,\"children\":[\"$\",\"div\",null,{\"className\":\"-mx-4 -mt-4 mb-4 border-b border-n700 p-4 md:-mx-8 md:p-8\",\"children\":[[\"$\",\"$Le1\",null,{\"quickStartCardProps\":{\"optionsData\":{\"options\":[{\"name\":\"llama-3_3-nemotron-49b-instruct\"},{\"name\":\"deepseek-r1\"},{\"name\":\"llama-3_3-70b-instruct\"},{\"name\":\"deepseek-r1-distill-llama-70b\"},{\"name\":\"embed-qa-4\"},{\"name\":\"qwen2_5-coder-32b-instruct\"}]},\"optionSpecs\":[null,{\"endpointSpec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for deepseek-ai/deepseek-r1\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"MIT License\",\"url\":\"https://huggingface.co/deepseek-ai/deepseek-r1/blob/main/LICENSE\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"deepseek-ai/deepseek-r1\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"user\",\"assistant\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.6,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":4096,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[\"$e2\"],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"}},\"text/event-stream\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[\"$e2\"],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":\"$eb\",\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Which number is larger, 9.11 or 9.8?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Which number is larger, 9.11 or 9.8?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The number 9.11 is...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"How many 'r's are in 'strawberry'?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"How many 'r's are in 'strawberry'?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 4096,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"In the word strawberry...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"$f4\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"$f5\",\"curl\":\"$f6\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"deepseek-ai/deepseek-r1\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":\"$eb\",\"ChatCompletion\":\"$f7\",\"ChatCompletionChunk\":\"$115\",\"ChatRequest\":\"$127\",\"Choice\":\"$fb\",\"ChoiceChunk\":\"$119\",\"Message\":\"$e2\",\"Usage\":\"$10b\"}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-15T15:30:42.018Z\",\"nvcfFunctionId\":\"854db4e5-9be7-45a0-a730-183cadf87e50\",\"createdDate\":\"2025-01-30T23:29:48.101Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-r1\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e NVIDIA Community Model License\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eMIT License\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"playground\":{\"type\":\"chatWithThinking\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$13b\"}]},\"artifactName\":\"deepseek-r1\"},\"artifact\":{\"artifact\":{\"name\":\"deepseek-r1\",\"displayName\":\"deepseek-r1\",\"publisher\":\"deepseek-ai\",\"shortDescription\":\"State-of-the-art, high-efficiency LLM excelling in reasoning, math, and coding.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/deepseek-r1.jpg\",\"labels\":[\"Math\",\"advanced reasoning\",\"chat\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2025-01-30T23:29:47.817Z\",\"description\":\"$13c\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-15T15:30:41.527Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"bb350840-a3f5-4a69-8438-a43f06ff0baa\"}},\"openApiSpec\":\"$13d\"},{\"endpointSpec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for meta/llama-3.3-70b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/meta-llama-3_3-70b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Llama 3.3 License\",\"url\":\"https://www.llama.com/llama3_3/license/\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"meta/llama-3.3-70b-instruct\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"system\",\"context\",\"user\",\"assistant\",\"tool\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}]},\"tool_call_id\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Tool Call Id\",\"description\":\"The id of the tool call.\"},\"tool_calls\":{\"anyOf\":[{\"items\":{\"properties\":{\"id\":{\"type\":\"string\",\"title\":\"Id\"},\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"},\"function\":{\"properties\":{\"name\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Name\"},\"arguments\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Arguments\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\",\"arguments\"],\"title\":\"FunctionCall\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ToolCall\"},\"type\":\"array\"},{\"type\":\"null\"}],\"title\":\"Tool Calls\",\"description\":\"The tool(s) called by the model.\"}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"minimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"tools\":{\"anyOf\":[{\"items\":{\"properties\":{\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"},\"function\":{\"properties\":{\"name\":{\"type\":\"string\",\"title\":\"Name\"},\"description\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Description\"},\"parameters\":{\"anyOf\":[{\"type\":\"object\"},{\"type\":\"null\"}],\"title\":\"Parameters\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\"],\"title\":\"FunctionDefinition\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ChatCompletionToolsParam\"},\"type\":\"array\"},{\"type\":\"null\"}],\"title\":\"Tools\"},\"frequency_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Frequency Penalty\",\"description\":\"Indicates how much to penalize new tokens based on their existing frequency in the text so far, decreasing model likelihood to repeat the same line verbatim.\"},\"presence_penalty\":{\"type\":\"number\",\"maximum\":2,\"minimum\":-2,\"default\":0,\"title\":\"Presence Penalty\",\"description\":\"Positive values penalize new tokens based on whether they appear in the text so far, increasing model likelihood to talk about new topics.\"},\"max_tokens\":{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4096,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[\"$172\"],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\",\"tool_calls\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"}},\"text/event-stream\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[\"$172\"],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\",\"tool_calls\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":\"$194\",\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"Tell me about Dumbledore.\",\"requestJson\":\"$19d\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"What is the weather in Santa Clara, CA?\",\"requestJson\":\"$19e\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"$19f\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n \\n\",\"node.js\":\"$1a0\",\"curl\":\"$1a1\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"meta/llama-3.3-70b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":\"$194\",\"ChatCompletion\":\"$1a2\",\"ChatCompletionChunk\":\"$1c0\",\"ChatRequest\":\"$1d2\",\"Choice\":\"$1a6\",\"ChoiceChunk\":\"$1c4\",\"Message\":\"$172\",\"ToolCall\":\"$181\",\"FunctionCall\":\"$186\",\"ChatCompletionToolsParam\":\"$1de\",\"ChatCompletionNamedFunction\":{\"properties\":{\"name\":{\"anyOf\":[{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Name\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"name\"],\"title\":\"ChatCompletionNamedFunction\"},\"ChatCompletionNamedToolChoiceParam\":{\"properties\":{\"function\":\"$1fb\",\"type\":{\"type\":\"string\",\"enum\":[\"function\"],\"const\":\"function\",\"title\":\"Type\",\"default\":\"function\"}},\"additionalProperties\":false,\"type\":\"object\",\"required\":[\"function\"],\"title\":\"ChatCompletionNamedToolChoiceParam\"},\"FunctionDefinition\":\"$1e2\",\"Usage\":\"$1b6\"}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-22T20:54:50.987Z\",\"nvcfFunctionId\":\"84eb5de1-166b-4bb4-a01b-4f51bd90aa52\",\"createdDate\":\"2024-12-06T17:04:43.514Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/meta-llama-3_3-70b-instruct\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: This trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-community-models-license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e NVIDIA Community Model License \u003c/a\u003e. ADDITIONAL INFORMATION: \u003ca href=\\\"https://www.llama.com/llama3_3/license/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e Llama 3.3 Community License Agreement\u003c/a\u003e, Built with Llama.\\n\",\"showUnavailableBanner\":false,\"cta\":{\"text\":\"Apply to Self-Host\",\"url\":\"https://www.nvidia.com/en-us/ai/nim-notifyme/\",\"nim_available_override_url\":\"https://catalog.ngc.nvidia.com/orgs/nim/teams/meta/containers/llama-3.3-70b-instruct\"},\"playground\":{\"type\":\"chatWithTools\"},\"deploy\":[{\"label\":\"Linux with Docker\",\"filename\":\"linux.md\",\"contents\":\"$202\"},{\"label\":\"Linux / Windows with NVIDIA AI Workbench\",\"filename\":\"workbench.md\",\"contents\":\"$203\"}]},\"artifactName\":\"llama-3_3-70b-instruct\"},\"artifact\":{\"artifact\":{\"name\":\"llama-3_3-70b-instruct\",\"displayName\":\"llama-3.3-70b-instruct\",\"publisher\":\"meta\",\"shortDescription\":\"Advanced LLM for reasoning, math, general knowledge, and function calling\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/llama-3_3-70b-instruct.jpg\",\"labels\":[\"Instruction following\",\"Math\",\"Reasoning\",\"Text-to-Text\",\"Code Generation\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"true\"},{\"key\":\"PREVIEW\",\"value\":\"false\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-12-06T17:04:43.078Z\",\"description\":\"$204\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-03-22T20:54:50.279Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"ae1856ed-0b8b-409d-a93a-9c4c03191681\"}},\"openApiSpec\":\"$205\"},null,{\"endpointSpec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for nvidia/embed-qa-4\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/nvidia-embed-qa-4 for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://nvidia.com/legal/terms-of-use\",\"contact\":{\"name\":\"NVIDIA Support\",\"url\":\"https://help.nvidia.com/\"},\"license\":{\"name\":\"CC-BY-NC-4.0\",\"url\":\"https://spdx.org/licenses/CC-BY-NC-4.0\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1\"}],\"paths\":{\"/embeddings\":{\"post\":{\"tags\":[\"Embeddings\"],\"summary\":\"Creates an embedding vector from the input text.\",\"operationId\":\"create_embedding\",\"requestBody\":{\"required\":true,\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"input\":{\"description\":\"Input text to embed. Max length depends on model.\",\"oneOf\":[{\"type\":\"string\"},{\"type\":\"array\",\"items\":{\"type\":\"string\"}}],\"minLength\":1,\"maxLength\":4096,\"title\":\"Input\"},\"model\":{\"type\":\"string\",\"description\":\"ID of the embedding model.\",\"example\":\"nvidia/embed-qa-4\",\"title\":\"Model\"},\"input_type\":{\"type\":\"string\",\"enum\":[\"passage\",\"query\"],\"description\":\"nvidia/embed-qa-4 and E5 models operate in `passage` or `query` mode, and thus require the `input_type` parameter. `passage` is used when generating embeddings during indexing. `query` is used when generating embeddings during querying. It is very important to use the correct `input_type`. Failure to do so will result in large drops in retrieval accuracy. As an alternative, it is possible to add the `-query` or `-passage` suffix to the `model` parameter like `nvidia/embed-qa-4-query` and not use the `input_type` field at all for OpenAI API compliance. Please note that the GTE model _does not_ accept the `input_type` parameter since both the query and passage are processed in the same way.\",\"title\":\"Input Type\"},\"encoding_format\":{\"type\":\"string\",\"description\":\"The format to return the embeddings in.\",\"enum\":[\"float\",\"base64\"],\"default\":\"float\",\"title\":\"Encoding Format\"},\"truncate\":{\"type\":\"string\",\"description\":\"Specifies how inputs longer than the maximum token length of the model are handled. Passing `START` discards the start of the input. `END` discards the end of the input. In both cases, input is discarded until the remaining input is exactly the maximum input token length for the model. If `NONE` is selected, when the input exceeds the maximum input token length an error will be returned.\",\"enum\":[\"NONE\",\"START\",\"END\"],\"default\":\"NONE\",\"title\":\"Truncate\"},\"user\":{\"type\":\"string\",\"description\":\"Not implemented, but provided for API compliance. This field is ignored.\",\"title\":\"User\"}},\"required\":[\"input\",\"model\"]}}}},\"x-nvai-meta\":{\"name\":\"Create Text Embedding\",\"description\":\"Generates an embedding vector from the provided text\\nusing a specified model. The embedding can be returned\\nin either float array or base64-encoded format.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Embedding vector for text input\",\"requestJson\":\"{\\n \\\"input\\\": \\\"What is the capital of France?\\\",\\n \\\"model\\\": \\\"nvidia/embed-qa-4\\\",\\n \\\"input_type\\\": \\\"query\\\",\\n \\\"encoding_format\\\": \\\"float\\\",\\n \\\"truncate\\\": \\\"NONE\\\"\\n}\\n\",\"responseJson\":\"$242\"}],\"templates\":[{\"title\":\"Synchronous requests\",\"requestEjs\":{\"curl\":\"curl -X POST https://integrate.api.nvidia.com/v1/embeddings \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"input\\\": [\\\"\u003c%- request.input.replaceAll('\\\"', '\\\\\\\\\\\"').replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e\\\"],\\n \\\"model\\\": \\\"nvidia/embed-qa-4\\\",\\n \\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\",\\n \\\"encoding_format\\\": \\\"\u003c%- request.encoding_format %\u003e\\\",\\n \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"\\n }'\\n\",\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n api_key=\\\"$NVIDIA_API_KEY\\\",\\n base_url=\\\"https://integrate.api.nvidia.com/v1\\\"\\n)\\n\\nresponse = client.embeddings.create(\\n input=[\u003c%- JSON.stringify(request.input) %\u003e],\\n model=\\\"nvidia/embed-qa-4\\\",\\n encoding_format=\\\"\u003c%- request.encoding_format %\u003e\\\",\\n extra_body={\\\"input_type\\\": \\\"\u003c%- request.input_type %\u003e\\\", \\\"truncate\\\": \\\"\u003c%- request.truncate %\u003e\\\"}\\n)\\n\\nprint(response.data[0].embedding)\\n\",\"langchain\":\"from langchain_nvidia_ai_endpoints import NVIDIAEmbeddings\\n\\nclient = NVIDIAEmbeddings(\\n model=\\\"NV-Embed-QA\\\", \\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n truncate=\\\"\u003c%- request.truncate %\u003e\\\", \\n )\\n\\nembedding = client.embed_query(\u003c%- JSON.stringify(request.input) %\u003e)\\nprint(embedding)\\n\"},\"response\":\"$243\"}]},\"responses\":{\"200\":{\"description\":\"Successful response\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\",\"example\":\"list\"},\"data\":{\"type\":\"array\",\"items\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\",\"example\":\"embedding\"},\"embedding\":{\"oneOf\":[{\"type\":\"array\",\"items\":{\"type\":\"number\"},\"description\":\"The embedding vector as a list of floats. The length of the vector depends on the model.\"},{\"type\":\"string\",\"description\":\"The embedding vector as a Base64 string. The length of the string depends on the model.\"}]},\"index\":{\"type\":\"integer\",\"description\":\"The index of the embedding in the list of embeddings.\"}},\"required\":[\"object\",\"embedding\",\"index\"]}},\"model\":{\"type\":\"string\",\"example\":\"nre-001\"},\"usage\":{\"type\":\"object\",\"description\":\"Not implemented, but provided for API compliance. This field will contain zeros.\",\"properties\":{\"prompt_tokens\":{\"type\":\"integer\",\"example\":0},\"total_tokens\":{\"type\":\"integer\",\"example\":0}}}}}}}},\"400\":{\"description\":\"Bad request\",\"content\":{\"application/json\":{\"schema\":{\"type\":\"object\",\"properties\":{\"object\":{\"type\":\"string\"},\"message\":{\"type\":\"string\"},\"detail\":{\"type\":\"object\",\"additionalProperties\":true},\"type\":{\"type\":\"string\"}}}}}}}}}},\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"EmbeddingObject\":\"$244\"}}},\"namespace\":\"qc69jvmznzxy\",\"updatedDate\":\"2025-01-17T20:00:27.769Z\",\"nvcfFunctionId\":\"09c64e32-2b65-4892-a285-2f585408d118\",\"createdDate\":\"2024-03-17T01:57:28.382Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/nvidia-embed-qa-4\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: Your use of this API is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Service Terms of Use\u003c/a\u003e; and the use of this model is governed by the \u003ca href=\\\"https://opensource.org/license/MIT\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eMIT License\u003c/a\u003e.\\n\",\"showUnavailableBanner\":false},\"artifactName\":\"embed-qa-4\"},\"artifact\":{\"artifact\":{\"name\":\"embed-qa-4\",\"displayName\":\"embed-qa-4\",\"publisher\":\"nvidia\",\"shortDescription\":\"GPU-accelerated generation of text embeddings used for question-answering retrieval.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/embed-qa-4.jpg\",\"labels\":[\"Embeddings\",\"Retrieval Augmented Generation\",\"Retrieval Augmented Generation\",\"Text-to-Embedding\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"bias\":\"Field | Response\\n:---------------------------------------------------------------------------------------------------|:---------------\\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None\\nMeasures taken to mitigate against unwanted bias: | None\",\"canGuestDownload\":true,\"createdDate\":\"2024-03-17T01:57:28.083Z\",\"description\":\"$24e\",\"explainability\":\"$24f\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"privacy\":\"$250\",\"safetyAndSecurity\":\"| Field | Response |\\n|:-------------------|:---------|\\n|Model Application(s): | Text Embedding for Retrieval |\\n|Describe the life-critical impact (if present). | Not Applicable |\\n|Use Case Restrictions:| Evaluation license for Non-Commerical Use Only. |\\n|Model and dataset restrictions:| The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. |\",\"updatedDate\":\"2025-01-17T20:00:27.231Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"170dc82a-30e6-4f87-9777-a0bd8ec530d8\"}},\"openApiSpec\":\"$251\"},{\"endpointSpec\":{\"openAPISpec\":{\"openapi\":\"3.1.0\",\"info\":{\"title\":\"NVIDIA NIM API for qwen/qwen2.5-coder-32b-instruct\",\"description\":\"The NVIDIA NIM REST API. Please see https://docs.api.nvidia.com/nim/reference/qwen-qwen2_5-coder-32b-instruct for more details.\",\"version\":\"1.0.0\",\"termsOfService\":\"https://www.nvidia.com/en-us/data-center/products/nvidia-ai-enterprise/eula/\",\"contact\":{\"name\":\"NVIDIA Enterprise Support\",\"url\":\"https://www.nvidia.com/en-us/support/enterprise/\"},\"license\":{\"name\":\"Apache 2.0\",\"url\":\"https://apache.org/licenses/LICENSE-2.0\"}},\"servers\":[{\"url\":\"https://integrate.api.nvidia.com/v1/\"}],\"paths\":{\"/chat/completions\":{\"post\":{\"operationId\":\"create_chat_completion_v1_chat_completions_post\",\"tags\":[\"Chat\"],\"summary\":\"Creates a model response for the given chat conversation.\",\"description\":\"Given a list of messages comprising a conversation, the model will return a response. Compatible with OpenAI. See https://platform.openai.com/docs/api-reference/chat/create\",\"requestBody\":{\"content\":{\"application/json\":{\"schema\":{\"additionalProperties\":false,\"properties\":{\"model\":{\"type\":\"string\",\"title\":\"Model\",\"default\":\"qwen/qwen2.5-coder-32b-instruct\"},\"messages\":{\"description\":\"A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present; `context` is also optional, but must come before a user question.\",\"examples\":[[{\"content\":\"I am going to Paris, what should I see?\",\"role\":\"user\"}]],\"items\":{\"additionalProperties\":false,\"properties\":{\"role\":{\"description\":\"The role of the message author.\",\"enum\":[\"system\",\"context\",\"user\",\"assistant\"],\"title\":\"Role\",\"type\":\"string\"},\"content\":{\"description\":\"The contents of the message.\",\"title\":\"Content\",\"type\":\"string\"}},\"required\":[\"role\",\"content\"],\"title\":\"Message\",\"type\":\"object\"},\"title\":\"Messages\",\"type\":\"array\"},\"temperature\":{\"default\":0.2,\"description\":\"The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Temperature\",\"type\":\"number\"},\"top_p\":{\"default\":0.7,\"description\":\"The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.\",\"maximum\":1,\"exclusiveMinimum\":0,\"title\":\"Top P\",\"type\":\"number\"},\"max_tokens\":{\"default\":1024,\"description\":\"The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.\",\"maximum\":4000,\"minimum\":1,\"title\":\"Max Tokens\",\"type\":\"integer\"},\"seed\":{\"anyOf\":[{\"maximum\":18446744073709552000,\"minimum\":0,\"type\":\"integer\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.\",\"examples\":[42],\"title\":\"Seed\"},\"stream\":{\"default\":false,\"description\":\"If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.\",\"title\":\"Stream\",\"type\":\"boolean\"},\"stop\":{\"anyOf\":[{\"items\":{\"type\":\"string\"},\"type\":\"array\"},{\"type\":\"string\"},{\"type\":\"null\"}],\"title\":\"Stop\",\"description\":\"A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.\"}},\"required\":[\"messages\"],\"title\":\"ChatRequest\",\"type\":\"object\"}}},\"required\":true},\"responses\":{\"200\":{\"description\":\"Invocation is fulfilled\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"message\":{\"allOf\":[\"$290\"],\"description\":\"A chat completion message generated by the model.\",\"examples\":[{\"content\":\"Ah, Paris, the City of Light! There are so many amazing things to see and do in this beautiful city ...\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.\",\"examples\":[\"stop\"],\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"message\"],\"title\":\"Choice\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"},\"usage\":{\"allOf\":[{\"properties\":{\"completion_tokens\":{\"description\":\"Number of tokens in the generated completion.\",\"examples\":[25],\"title\":\"Completion Tokens\",\"type\":\"integer\"},\"prompt_tokens\":{\"description\":\"Number of tokens in the prompt.\",\"examples\":[9],\"title\":\"Prompt Tokens\",\"type\":\"integer\"},\"total_tokens\":{\"description\":\"Total number of tokens used in the request (prompt + completion).\",\"examples\":[34],\"title\":\"Total Tokens\",\"type\":\"integer\"}},\"required\":[\"completion_tokens\",\"prompt_tokens\",\"total_tokens\"],\"title\":\"Usage\",\"type\":\"object\"}],\"description\":\"Usage statistics for the completion request.\"}},\"required\":[\"id\",\"choices\",\"usage\"],\"title\":\"ChatCompletion\",\"type\":\"object\"}},\"text/event-stream\":{\"schema\":{\"properties\":{\"id\":{\"description\":\"A unique identifier for the completion.\",\"format\":\"uuid\",\"title\":\"Id\",\"type\":\"string\"},\"choices\":{\"description\":\"The list of completion choices the model generated for the input prompt.\",\"items\":{\"properties\":{\"index\":{\"description\":\"The index of the choice in the list of choices (always 0).\",\"title\":\"Index\",\"type\":\"integer\"},\"delta\":{\"allOf\":[\"$290\"],\"description\":\"A chat completion delta generated by streamed model responses.\",\"examples\":[{\"content\":\"Ah,\",\"role\":\"assistant\"}]},\"finish_reason\":{\"anyOf\":[{\"enum\":[\"stop\",\"length\"],\"type\":\"string\"},{\"type\":\"null\"}],\"default\":null,\"description\":\"The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.\",\"title\":\"Finish Reason\"}},\"required\":[\"index\",\"delta\"],\"title\":\"ChoiceChunk\",\"type\":\"object\"},\"title\":\"Choices\",\"type\":\"array\"}},\"required\":[\"id\",\"choices\"],\"title\":\"ChatCompletionChunk\",\"type\":\"object\"}}}},\"202\":{\"description\":\"Result is pending. Client should poll using the requestId.\\n\",\"content\":{\"application/json\":{\"example\":{},\"schema\":{}}},\"headers\":{\"NVCF-REQID\":{\"description\":\"requestId required for pooling\",\"schema\":{\"type\":\"string\",\"format\":\"uuid\"}},\"NVCF-STATUS\":{\"description\":\"Invocation status\",\"schema\":{\"type\":\"string\"}}}},\"422\":{\"description\":\"Validation failed, provided entity could not be processed.\",\"content\":{\"application/json\":{\"schema\":{\"properties\":{\"type\":{\"type\":\"string\",\"description\":\"Error type\"},\"title\":{\"type\":\"string\",\"description\":\"Error title\"},\"status\":{\"type\":\"integer\",\"description\":\"Error status code\"},\"detail\":{\"type\":\"string\",\"description\":\"Detailed information about the error\"},\"instance\":{\"type\":\"string\",\"description\":\"Function instance used to invoke the request\"},\"requestId\":{\"type\":\"string\",\"format\":\"uuid\",\"description\":\"UUID of the request\"}},\"type\":\"object\",\"required\":[\"type\",\"title\",\"status\",\"detail\",\"instance\",\"requestId\"],\"title\":\"InvokeError\"},\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:unprocessable-entity\",\"title\":\"Unprocessable Entity\",\"status\":422,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}},\"500\":{\"description\":\"The invocation ended with an error.\",\"content\":{\"application/json\":{\"schema\":\"$296\",\"example\":{\"type\":\"urn:nvcf-worker-service:problem-details:internal-server-error\",\"title\":\"Internal Server Error\",\"status\":500,\"detail\":\"string\",\"instance\":\"/v2/nvcf/pexec/functions/4a58c6cb-a9b4-4014-99de-3e704d4ae687\",\"requestId\":\"3fa85f64-5717-4562-b3fc-2c963f66afa6\"}}}}},\"x-nvai-meta\":{\"name\":\"Create chat completion\",\"returns\":\"Returns a [chat completion](/docs/api-reference/chat/object) object, or a streamed sequence of [chat completion chunk](/docs/api-reference/chat/streaming) objects if the request is streamed.\\n\",\"path\":\"create\",\"examples\":[{\"name\":\"Write a limerick about the wonders of GPU computing.\",\"requestJson\":\"{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"Write a limerick about the wonders of GPU computing.\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"The python functions...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"},{\"name\":\"What can I see at NVIDIA's GPU Technology Conference?\",\"requestJson\":\"{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": [\\n {\\n \\\"role\\\": \\\"user\\\",\\n \\\"content\\\": \\\"What can I see at NVIDIA's GPU Technology Conference?\\\"\\n }\\n ],\\n \\\"top_p\\\": 0.7,\\n \\\"max_tokens\\\": 1024,\\n \\\"seed\\\": 42,\\n \\\"stream\\\": true\\n}\\n\",\"responseJson\":\"{\\n \\\"id\\\": \\\"id-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"At NVIDIA's GPU Technology Conference (GTC)...\\\"\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}],\"templates\":[{\"title\":\"No Streaming\",\"requestEjs\":{\"python\":\"from openai import OpenAI\\n\\nclient = OpenAI(\\n base_url = \\\"https://integrate.api.nvidia.com/v1\\\",\\n api_key = \\\"$NVIDIA_API_KEY\\\"\\n)\\n\\ncompletion = client.chat.completions.create(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n messages=\u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n stream=\u003c%- request.stream?.toString()[0].toUpperCase() + request.stream?.toString().slice(1) %\u003e\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in completion:\\n if chunk.choices[0].delta.content is not None:\\n print(chunk.choices[0].delta.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nprint(completion.choices[0].message)\\n\u003c% } %\u003e\\n\",\"langChain\":\"from langchain_nvidia_ai_endpoints import ChatNVIDIA\\n\\nclient = ChatNVIDIA(\\n model=\\\"\u003c%- request.model %\u003e\\\",\\n api_key=\\\"$NVIDIA_API_KEY\\\", \\n temperature=\u003c%- request.temperature %\u003e,\\n top_p=\u003c%- request.top_p %\u003e,\\n max_tokens=\u003c%- request.max_tokens %\u003e,\\n)\\n\u003c% if (request.stream) { %\u003e\\nfor chunk in client.stream(\u003c%- JSON.stringify(request.messages) %\u003e): \\n print(chunk.content, end=\\\"\\\")\\n\u003c% } else { %\u003e\\nresponse = client.invoke(\u003c%- JSON.stringify(request.messages) %\u003e)\\nprint(response.content)\\n\u003c% } %\u003e\\n\",\"node.js\":\"import OpenAI from 'openai';\\n\\nconst openai = new OpenAI({\\n apiKey: '$NVIDIA_API_KEY',\\n baseURL: 'https://integrate.api.nvidia.com/v1',\\n})\\n\\nasync function main() {\\n const completion = await openai.chat.completions.create({\\n model: \\\"\u003c%- request.model %\u003e\\\",\\n messages: \u003c%- JSON.stringify(request.messages) %\u003e,\\n temperature: \u003c%- request.temperature %\u003e,\\n top_p: \u003c%- request.top_p %\u003e,\\n max_tokens: \u003c%- request.max_tokens %\u003e,\\n stream: \u003c%- request.stream %\u003e,\\n })\\n \u003c% if (request.stream) { %\u003e\\n for await (const chunk of completion) {\\n process.stdout.write(chunk.choices[0]?.delta?.content || '')\\n }\\n \u003c% } else { %\u003e\\n process.stdout.write(completion.choices[0]?.message?.content);\\n \u003c% } %\u003e\\n}\\n\\nmain();\",\"curl\":\"curl https://integrate.api.nvidia.com/v1/chat/completions \\\\\\n -H \\\"Content-Type: application/json\\\" \\\\\\n -H \\\"Authorization: Bearer $NVIDIA_API_KEY\\\" \\\\\\n -d '{\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"messages\\\": \u003c%- JSON.stringify(request.messages).replaceAll(\\\"'\\\", \\\"'\\\\\\\"'\\\\\\\"'\\\") %\u003e,\\n \\\"temperature\\\": \u003c%- request.temperature %\u003e, \\n \\\"top_p\\\": \u003c%- request.top_p %\u003e,\\n \\\"max_tokens\\\": \u003c%- request.max_tokens %\u003e,\\n \\\"stream\\\": \u003c%- request.stream %\u003e \\n }'\\n\"},\"response\":\"{\\n \\\"id\\\": \\\"chatcmpl-123\\\",\\n \\\"object\\\": \\\"chat.completion\\\",\\n \\\"created\\\": 1677652288,\\n \\\"model\\\": \\\"qwen/qwen2.5-coder-32b-instruct\\\",\\n \\\"system_fingerprint\\\": \\\"fp_44709d6fcb\\\",\\n \\\"choices\\\": [{\\n \\\"index\\\": 0,\\n \\\"message\\\": {\\n \\\"role\\\": \\\"assistant\\\",\\n \\\"content\\\": \\\"\\\\n\\\\nHello there, how may I assist you today?\\\",\\n },\\n \\\"finish_reason\\\": \\\"stop\\\"\\n }],\\n \\\"usage\\\": {\\n \\\"prompt_tokens\\\": 9,\\n \\\"completion_tokens\\\": 12,\\n \\\"total_tokens\\\": 21\\n }\\n}\\n\"}]}}}},\"security\":[{\"Token\":[]}],\"components\":{\"securitySchemes\":{\"Token\":{\"type\":\"http\",\"scheme\":\"bearer\"}},\"schemas\":{\"Errors\":\"$296\",\"ChatCompletion\":\"$29f\",\"ChatCompletionChunk\":\"$2bd\",\"ChatRequest\":\"$2cf\",\"Choice\":\"$2a3\",\"ChoiceChunk\":\"$2c1\",\"Message\":\"$290\",\"Usage\":\"$2b3\"}}},\"namespace\":\"qc69jvmznzxy\",\"nvcfFunctionId\":\"f3160eb9-d42a-46a6-b7da-aad54d4b2134\",\"createdDate\":\"2024-12-20T16:43:06.694Z\",\"attributes\":{\"apiDocsUrl\":\"https://docs.api.nvidia.com/nim/reference/qwen-qwen2_5-coder-32b-instruct\",\"termsOfUse\":\"\u003cb\u003eGOVERNING TERMS\u003c/b\u003e: The trial service is governed by the \u003ca href=\\\"https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eNVIDIA API Trial Terms of Service\u003c/a\u003e. Use of this model is governed by the \u003ca href=\\\"https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-software-license-agreement/\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003e NVIDIA Open Model License Agreement\u003c/a\u003e. Additional Information: \u003ca href=\\\"https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct/blob/main/LICENSE\\\" rel=\\\"noreferrer\\\" target=\\\"_blank\\\"\u003eApache License Version 2.0\u003c/a\u003e\\n\",\"showUnavailableBanner\":false,\"playground\":{\"type\":\"chat\"}},\"artifactName\":\"qwen2_5-coder-32b-instruct\"},\"artifact\":{\"artifact\":{\"name\":\"qwen2_5-coder-32b-instruct\",\"displayName\":\"qwen2.5-coder-32b-instruct\",\"publisher\":\"qwen\",\"shortDescription\":\"Advanced LLM for code generation, reasoning, and fixing across popular programming languages.\",\"logo\":\"https://assets.ngc.nvidia.com/products/api-catalog/images/qwen2_5-coder-32b-instruct.jpg\",\"labels\":[\"code completion\",\"code generation\",\"text-to-code\"],\"attributes\":[{\"key\":\"AVAILABLE\",\"value\":\"false\"},{\"key\":\"PREVIEW\",\"value\":\"true\"}],\"artifactType\":\"ENDPOINT\",\"canGuestDownload\":true,\"createdDate\":\"2024-12-20T16:43:06.305Z\",\"description\":\"$2e6\",\"isPublic\":true,\"isReadOnly\":true,\"orgName\":\"qc69jvmznzxy\",\"updatedDate\":\"2024-12-20T16:43:06.305Z\"},\"requestStatus\":{\"statusCode\":\"SUCCESS\",\"requestId\":\"08ab5aa0-e3a7-4ebf-89b7-6f79a1fc9420\"}},\"openApiSpec\":\"$2e7\"}]}}],[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"top\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\",\"explore\",\"top\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"notFoundStyles\":\"$undefined\"}]]}]}],[\"$\",\"div\",null,{\"className\":\"-mx-4 -mt-4 p-4 md:-mx-8 md:p-8\",\"children\":[\"$\",\"div\",null,{\"className\":\"mx-auto flex max-w-[1600px]\",\"children\":[[\"$\",\"div\",null,{\"className\":\"hidden w-[220px] md:block\",\"children\":[\"$\",\"$L319\",null,{\"initialData\":{\"models\":[{\"title\":\"Reasoning\",\"config\":\"/reasoning\"},{\"title\":\"Vision\",\"config\":\"/vision\"},{\"title\":\"Visual Design\",\"config\":\"/visual-design\"},{\"title\":\"Retrieval\",\"config\":\"/retrieval\"},{\"title\":\"Speech\",\"config\":\"/speech\"},{\"title\":\"Biology\",\"config\":\"/biology\"},{\"title\":\"Simulation\",\"config\":\"/simulation\"},{\"title\":\"Climate \u0026 Weather\",\"config\":\"/climate-weather\"},{\"title\":\"Safety \u0026 Moderation\",\"config\":\"/safety-moderation\"}],\"industries\":[{\"title\":\"Automotive\",\"config\":\"/automotive\"},{\"title\":\"Gaming\",\"config\":\"/gaming\"},{\"title\":\"Healthcare\",\"config\":\"/healthcare\"},{\"title\":\"Industrial\",\"config\":\"/industrial\"},{\"title\":\"Robotics\",\"config\":\"/robotics\"}]}}]}],[\"$\",\"div\",null,{\"className\":\"flex w-full min-w-0 flex-1 flex-col border-n700 md:border-l md:pl-8\",\"children\":[\"$\",\"$L15\",null,{\"parallelRouterKey\":\"children\",\"segmentPath\":[\"children\",\"(app-bar-layout)\",\"children\",\"explore\",\"children\"],\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L16\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"notFoundStyles\":\"$undefined\"}]}]]}]}]]\n"])</script></body></html>