{"id":1391,"date":"2026-01-26T10:00:00","date_gmt":"2026-01-26T02:00:00","guid":{"rendered":"https:\/\/smartaiunion.com\/?p=1391"},"modified":"2026-01-25T18:11:43","modified_gmt":"2026-01-25T10:11:43","slug":"vllm%e5%9b%a2%e9%98%9f%e5%ae%98%e5%ae%a3%e5%88%9b%e4%b8%9a%ef%bc%9a%e8%9e%8d%e8%b5%841-5%e4%ba%bf%e7%be%8e%e5%85%83%ef%bc%8c%e6%b8%85%e5%8d%8e%e7%89%b9%e5%a5%96%e6%b8%b8%e5%87%af%e8%b6%85%e6%88%90","status":"publish","type":"post","link":"https:\/\/smartaiunion.com\/index.php\/2026\/01\/26\/1391\/","title":{"rendered":"vLLM\u56e2\u961f\u5b98\u5ba3\u521b\u4e1a\uff1a\u878d\u8d441.5\u4ebf\u7f8e\u5143\uff0c\u6e05\u534e\u7279\u5956\u6e38\u51ef\u8d85\u6210\u4e3a\u8054\u521b"},"content":{"rendered":"\n<p>\u6709\u53f2\u4ee5\u6765\u89c4\u6a21\u6700\u5927\u7684\u79cd\u5b50\u8f6e\u878d\u8d44\u4e4b\u4e00\u3002<\/p>\n\n\n\n<p>\u5927\u6a21\u578b\u63a8\u7406\u7684\u57fa\u77f3 vLLM\uff0c\u73b0\u5728\u6210\u4e3a\u521b\u4e1a\u516c\u53f8\u4e86\u3002<\/p>\n\n\n\n<p>\u5317\u4eac\u65f6\u95f4\u5468\u4e94\u51cc\u6668\u4f20\u6765\u6d88\u606f\uff0c\u7531\u5f00\u6e90\u8f6f\u4ef6 vLLM \u7684\u521b\u5efa\u8005\u521b\u7acb\u7684\u4eba\u5de5\u667a\u80fd\u521d\u521b\u516c\u53f8 Inferact \u6b63\u5f0f\u6210\u7acb\uff0c\u5176\u5728\u79cd\u5b50\u8f6e\u878d\u8d44\u4e2d\u7b79\u96c6\u4e86 1.5 \u4ebf\u7f8e\u5143\uff08\u7ea6\u5408 10 \u4ebf\u5143\u4eba\u6c11\u5e01\uff09\uff0c\u516c\u53f8\u4f30\u503c\u8fbe\u5230 8 \u4ebf\u7f8e\u5143\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/img.36krcdn.com\/hsossms\/20260123\/v2_6ccb2d4eeec44252961c75a14759ad57@000000_oswg370898oswg1080oswg1111_img_000?x-oss-process=image\/format,jpg\/interlace,1\" alt=\"\"\/><\/figure>\n\n\n\n<p>\u672c\u8f6e\u878d\u8d44\u7531\u98ce\u9669\u6295\u8d44\u516c\u53f8 Andreessen Horowitz\uff08a16z\uff09\u548c Lightspeed \u9886\u6295\uff0cSequoia Capital\u3001Altimeter Capital\u3001Redpoint Ventures \u548c ZhenFund \u4e5f\u53c2\u4e0e\u4e86\u6295\u8d44\u3002<\/p>\n\n\n\n<p>Inferact \u7684 1.5 \u4ebf\u7f8e\u5143\u5929\u4f7f\u8f6e\u878d\u8d44\u867d\u4e0d\u53ca Ilya Sutskever \u7684\u516c\u53f8 SSI \u7684 10 \u4ebf\u7f8e\u5143\uff0c\u4f46\u5df2\u7ecf\u8d85\u8fc7\u4e86 Mistral AI \u7684 1.15 \u4ebf\u7f8e\u5143\uff0c\u662f\u6709\u53f2\u4ee5\u6765\u89c4\u6a21\u6700\u5927\u7684\u79cd\u5b50\u8f6e\u878d\u8d44\u4e4b\u4e00\uff0c\u6807\u5fd7\u7740\u4e1a\u754c\u5bf9\u4e8e AI \u63a8\u7406\u57fa\u7840\u8bbe\u65bd\u7684\u91cd\u89c6\u7a0b\u5ea6\u6b63\u5728\u6025\u901f\u63d0\u5347\u3002<\/p>\n\n\n\n<p>Inferact \u7684\u4f7f\u547d\u662f\u5c06 vLLM \u53d1\u5c55\u6210\u4e3a\u4e16\u754c\u9886\u5148\u7684 AI \u63a8\u7406\u5f15\u64ce\uff0c\u5e76\u901a\u8fc7\u964d\u4f4e\u63a8\u7406\u6210\u672c\u3001\u52a0\u5feb\u63a8\u7406\u901f\u5ea6\u6765\u52a0\u901f AI \u7684\u53d1\u5c55\u3002<\/p>\n\n\n\n<p>\u8be5\u516c\u53f8\u8ba4\u4e3a\uff0cAI \u884c\u4e1a\u672a\u6765\u9762\u4e34\u7684\u6700\u5927\u6311\u6218\u4e0d\u662f\u6784\u5efa\u65b0\u6a21\u578b\uff0c\u800c\u662f\u5982\u4f55\u4ee5\u4f4e\u6210\u672c\u3001\u9ad8\u53ef\u9760\u6027\u5730\u8fd0\u884c\u73b0\u6709\u6a21\u578b\u3002<\/p>\n\n\n\n<p>\u6beb\u65e0\u7591\u95ee\uff0cInferact \u7684\u6838\u5fc3\u662f\u5f00\u6e90\u9879\u76ee vLLM\uff0c\u8fd9\u662f\u4e00\u4e2a\u4e8e 2023 \u5e74\u542f\u52a8\u7684\u5f00\u6e90\u9879\u76ee\uff0c\u65e8\u5728\u5e2e\u52a9\u4f01\u4e1a\u5728\u6570\u636e\u4e2d\u5fc3\u786c\u4ef6\u4e0a\u9ad8\u6548\u8fd0\u884c AI \u6a21\u578b\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/img.36krcdn.com\/hsossms\/20260123\/v2_c6b944f21e124438ab8120f368127685@000000_oswg146842oswg1080oswg581_img_000?x-oss-process=image\/format,jpg\/interlace,1\" alt=\"\"\/><\/figure>\n\n\n\n<p>vLLM \u6700\u521d\u7531\u52a0\u5dde\u5927\u5b66\u4f2f\u514b\u5229\u5206\u6821\uff08UC Berkeley\uff09\u7684\u5929\u7a7a\u8ba1\u7b97\u5b9e\u9a8c\u5ba4 (Sky Computing Lab) \u5f00\u53d1\uff0c\u73b0\u7531 PyTorch \u57fa\u91d1\u4f1a\u8d1f\u8d23\u7ba1\u7406\uff0c\u5df2\u5438\u5f15\u4e86\u6765\u81ea\u6574\u4e2a AI \u884c\u4e1a\u7684 2000 \u591a\u540d\u8d21\u732e\u8005\uff0c\u662f\u5168\u7403\u8303\u56f4\u5185\u6700\u53d7\u6b22\u8fce\u7684\u5f00\u6e90\u5927\u6a21\u578b\u63a8\u7406\u52a0\u901f\u6846\u67b6\u3002<\/p>\n\n\n\n<p>\u5982\u4eca\uff0cvLLM \u7684\u63a8\u7406\u80fd\u529b\u5728\u4e3a Meta\u3001\u8c37\u6b4c\u3001Character.AI \u7b49\u79d1\u6280\u516c\u53f8\u63d0\u4f9b\u652f\u6301\u3002<\/p>\n\n\n\n<p>Inferact \u7684\u9996\u5e2d\u6267\u884c\u5b98 Simon Mo \u662f\u4e00\u4f4d\u4f2f\u514b\u5229\u5728\u8bfb\u535a\u58eb\u751f\uff0c\u4ed6\u662f vLLM \u7684\u521b\u59cb\u7ef4\u62a4\u8005\u4e4b\u4e00\u3002Mo \u8868\u793a\uff0c\u516c\u53f8\u6210\u7acb\u4e8e 2025 \u5e74 11 \u6708\uff0c\u5e76\u4e8e\u672c\u5468\u6b63\u5f0f\u5bf9\u5916\u516c\u5e03\u3002\u4ed6\u5c06 Inferact \u7684\u8d77\u6e90\u4e0e\u4f2f\u514b\u5229\u65e9\u671f\u7684\u4e00\u4e9b\u8f6f\u4ef6\u9879\u76ee\u8fdb\u884c\u4e86\u6bd4\u8f83\uff0c\u8fd9\u4e9b\u9879\u76ee\u540e\u6765\u53d1\u5c55\u6210\u4e3a\u89c4\u6a21\u66f4\u5927\u7684\u4f01\u4e1a\uff0c\u4f8b\u5982 Apache Spark \u548c Ray\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/img.36krcdn.com\/hsossms\/20260123\/v2_396bdd32baf34978a1009d6cbb290670@000000_oswg212929oswg460oswg460_img_000?x-oss-process=image\/format,jpg\/interlace,1\" alt=\"\"\/><\/figure>\n\n\n\n<p>\u5728\u5ba3\u5e03\u878d\u8d44\u7684\u540c\u65f6\uff0cLightspeed\uff08\u5149\u901f\u521b\u6295\uff09\u4e5f\u53d1\u5e03\u4e86\u5bf9 Simon Mo \u7684\u8bbf\u8c08\u3002\u5728\u5176\u4e2d Simon Mo \u8c08\u5230\u4e86\u5bf9\u4e8e\u5168\u7403 AI \u7b97\u529b\u7d27\u7f3a\u7684\u62c5\u5fe7\uff0c\u300c\u5f53\u524d\u7528\u4e8e\u5927\u6a21\u578b\u8bad\u7ec3\u7684 AI \u96c6\u7fa4\uff0c\u5c06\u5728\u516d\u4e2a\u6708\u5185\u5b8c\u5168\u88ab\u7528\u4e8e\u63a8\u7406\u2026\u2026 \u63a8\u7406\u4f1a\u9010\u6e10\u6d88\u8017\u6389\u6240\u6709\u7b97\u529b\u5bb9\u91cf\uff0c\u5e76\u8017\u5c3d\u6240\u6709\u65b0\u589e\u7684\u5bb9\u91cf\u3002\u300d<\/p>\n\n\n\n<p>\u5728\u516c\u544a\u4e2d\uff0cInferact \u8868\u793a\uff0c\u5176\u5b9a\u4f4d\u4e8e\u6a21\u578b\u548c\u786c\u4ef6\u7684\u4ea4\u6c47\u70b9\uff1a\u5f53\u6a21\u578b\u5382\u5546\u53d1\u5e03\u65b0\u67b6\u6784\u65f6\uff0c\u4ed6\u4eec\u4f1a\u4e0e vLLM \u5408\u4f5c\uff0c\u786e\u4fdd\u63d0\u4f9b\u9996\u65e5\u652f\u6301\uff1b\u5f53\u786c\u4ef6\u5382\u5546\u5f00\u53d1\u65b0\u82af\u7247\u65f6\uff0c\u4ed6\u4eec\u4f1a\u4e0e vLLM \u96c6\u6210\uff1b\u5f53\u5927\u6a21\u578b\u56e2\u961f\u8fdb\u884c\u5927\u89c4\u6a21\u90e8\u7f72\u65f6\uff0c\u4ed6\u4eec\u4f1a\u8fd0\u884c vLLM\uff0c\u4ece\u524d\u6cbf\u5b9e\u9a8c\u5ba4\u5230\u8d85\u5927\u89c4\u6a21\u6570\u636e\u4e2d\u5fc3\uff0c\u518d\u5230\u670d\u52a1\u6570\u767e\u4e07\u7528\u6237\u7684\u521d\u521b\u516c\u53f8\uff0c\u65e0\u4e00\u4f8b\u5916\u3002<\/p>\n\n\n\n<p>\u5982\u4eca\uff0cvLLM \u652f\u6301\u4e86 500 \u591a\u79cd\u6a21\u578b\u67b6\u6784\uff0c\u53ef\u5728 200 \u591a\u79cd\u52a0\u901f\u5668\u4e0a\u8fd0\u884c\uff0c\u5e76\u652f\u6301\u7740\u5168\u7403\u89c4\u6a21\u7684\u63a8\u7406\u3002\u8fd9\u4e2a\u7531 2000 \u591a\u4f4d\u8d21\u732e\u8005\u5171\u540c\u6784\u5efa\u7684\u751f\u6001\u7cfb\u7edf\uff0c\u662f Inferact \u5f97\u4ee5\u6210\u7acb\u7684\u57fa\u7840\u3002<\/p>\n\n\n\n<p><strong>Inferact \u8868\u793a\uff0c\u5176\u9996\u8981\u4efb\u52a1\u662f\u7ee7\u7eed\u652f\u6301 vLLM \u4f5c\u4e3a\u72ec\u7acb\u7684\u5f00\u6e90\u9879\u76ee\uff0c\u5e76\u5c06\u6539\u8fdb\u6210\u679c\u5206\u4eab\u7ed9\u793e\u533a\u3002<\/strong>\u4ed6\u4eec\u8ba1\u5212\u8fdb\u4e00\u6b65\u63d0\u5347 vLLM \u7684\u6027\u80fd\uff0c\u6df1\u5316\u5bf9\u65b0\u5174\u6a21\u578b\u67b6\u6784\u7684\u652f\u6301\uff0c\u5e76\u6269\u5927\u5bf9\u524d\u6cbf\u786c\u4ef6\u7684\u8986\u76d6\u8303\u56f4\u3002Inferact \u7684\u7b2c\u4e8c\u4e2a\u76ee\u6807\u662f\u5f00\u53d1\u4e00\u6b3e\u72ec\u7acb\u7684\u5546\u4e1a\u4ea7\u54c1\uff0c\u5e2e\u52a9\u4f01\u4e1a\u5728\u4e0d\u540c\u7c7b\u578b\u7684\u786c\u4ef6\u4e0a\u66f4\u9ad8\u6548\u5730\u8fd0\u884c AI \u6a21\u578b\u3002<\/p>\n\n\n\n<p>\u503c\u5f97\u5173\u6ce8\u7684\u662f\uff0cvLLM \u9879\u76ee\u7684\u6838\u5fc3\u8d21\u732e\u8005\u6e05\u534e\u535a\u58eb\u6e38\u51ef\u8d85\u6210\u4e3a\u4e86\u8fd9\u5bb6\u516c\u53f8\u7684\u8054\u5408\u521b\u59cb\u4eba\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img decoding=\"async\" src=\"https:\/\/img.36krcdn.com\/hsossms\/20260123\/v2_6c98fb5c49324aa0a8973afd26d17264@000000_oswg98678oswg876oswg399_img_000?x-oss-process=image\/format,jpg\/interlace,1\" alt=\"\"\/><\/figure>\n\n\n\n<p>\u636e\u4ecb\u7ecd\uff0cInferact \u7684\u521b\u59cb\u56e2\u961f\u5305\u62ec Simon Mo\u3001Woosuk Kwon\u3001Kaichao You\uff08\u6e38\u51ef\u8d85\uff09\u3001Roger Wang\u3001Joseph Gonzalez\u3001Ion Stoica \u7b49\u4eba\u3002<\/p>\n\n\n\n<p>\u539f\u6587\u94fe\u63a5\uff1a<a href=\"https:\/\/36kr.com\/p\/3651294559003015\">https:\/\/36kr.com\/p\/3651294559003015<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u6709\u53f2\u4ee5\u6765\u89c4\u6a21\u6700\u5927\u7684\u79cd\u5b50\u8f6e\u878d\u8d44\u4e4b\u4e00\u3002 \u5927\u6a21\u578b\u63a8\u7406\u7684\u57fa\u77f3 vLLM\uff0c\u73b0\u5728&#46;&#46;&#46;<\/p>\n","protected":false},"author":2,"featured_media":671,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[],"class_list":["post-1391","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news"],"_links":{"self":[{"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/posts\/1391","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/comments?post=1391"}],"version-history":[{"count":1,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/posts\/1391\/revisions"}],"predecessor-version":[{"id":1392,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/posts\/1391\/revisions\/1392"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/media\/671"}],"wp:attachment":[{"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/media?parent=1391"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/categories?post=1391"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/smartaiunion.com\/index.php\/wp-json\/wp\/v2\/tags?post=1391"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}