{"id":55836,"date":"2025-08-14T08:54:44","date_gmt":"2025-08-14T00:54:44","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/55836.html"},"modified":"2025-08-14T08:54:44","modified_gmt":"2025-08-14T00:54:44","slug":"%e8%87%aa%e5%8a%a8%e8%b0%83%e4%bc%98-vllm-%e6%9c%8d%e5%8a%a1%e5%99%a8%e5%8f%82%e6%95%b0%ef%bc%88%e5%ae%9e%e6%88%98%e6%8c%87%e5%8d%97%ef%bc%89","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/55836.html","title":{"rendered":"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09"},"content":{"rendered":"<h3>1. \u64cd\u4f5c\u6b65\u9aa4<\/h3>\n<ul>\n<li>\u8be5\u811a\u672c\u80fd\u591f\u81ea\u52a8\u641c\u7d22\u6700\u4f18\u7684vLLM\u670d\u52a1\u5668\u53c2\u6570\u7ec4\u5408&#xff08;\u5305\u62ecmax-num-seqs\u548cmax-num-batched-tokens&#xff09;&#xff0c;\u5728\u6ee1\u8db3\u7aef\u5230\u7aef\u5ef6\u8fdf\u548c\u524d\u7f00\u7f13\u5b58\u547d\u4e2d\u7387\u7b49\u8981\u6c42\u7684\u540c\u65f6&#xff0c;\u5b9e\u73b0\u541e\u5410\u91cf\u6700\u5927\u5316\u3002<\/li>\n<\/ul>\n<h4>1.1 \u524d\u63d0\u6761\u4ef6<\/h4>\n<li>\u514b\u9686 vLLM \u5e76\u5207\u5230\u76ee\u6807\u5206\u652f<span class=\"token function\">git<\/span> clone https:\/\/github.com\/vllm-project\/vllm.git<br \/>\n<span class=\"token builtin class-name\">cd<\/span> vllm<br \/>\n<span class=\"token comment\"># git checkout &lt;your-branch&gt;<\/span><\/p>\n<\/li>\n<li>\u5b89\u88c5\u8fd0\u884c\u73af\u5883<br \/>\n\u5982\u679c\u4f7f\u7528 TPU&#xff0c;\u8bf7\u6fc0\u6d3b\u5bf9\u5e94 conda \u73af\u5883\u5e76\u5b89\u88c5\u5339\u914d\u7248\u672c\u7684 torch\u3001torch_xla\u3002<\/li>\n<li>\u6a21\u578b\u51c6\u5907<br \/>\n\u82e5\u4f7f\u7528\u81ea\u5b9a\u4e49\u6a21\u578b&#xff0c;\u786e\u4fdd\u914d\u7f6e\u6587\u4ef6\u653e\u7f6e\u6b63\u786e\u4e14\u53ef\u8bbf\u95ee\u3002<\/li>\n<h4>1.2 \u914d\u7f6e&#xff08;\u811a\u672c\u9876\u90e8\u5fc5\u987b\u8bbe\u7f6e&#xff09;<\/h4>\n<table>\n<tr>\u53d8\u91cf\u8bf4\u660e\u793a\u4f8b<\/tr>\n<tbody>\n<tr>\n<td>BASE<\/td>\n<td>vLLM \u4ed3\u5e93\u6240\u5728\u76ee\u5f55\u7684\u7edd\u5bf9\u8def\u5f84<\/td>\n<td>&#034;$HOME&#034;<\/td>\n<\/tr>\n<tr>\n<td>MODEL<\/td>\n<td>Hugging Face \u6a21\u578b\u540d\u79f0<\/td>\n<td>&#034;meta-llama\/Llama-3.1-8B-Instruct&#034;<\/td>\n<\/tr>\n<tr>\n<td>SYSTEM<\/td>\n<td>\u786c\u4ef6\u7c7b\u578b&#xff1a;TPU \u6216 GPU<\/td>\n<td>&#034;TPU&#034;<\/td>\n<\/tr>\n<tr>\n<td>TP<\/td>\n<td>Tensor-parallelism \u5927\u5c0f<\/td>\n<td>1<\/td>\n<\/tr>\n<tr>\n<td>DOWNLOAD_DIR<\/td>\n<td>\u6a21\u578b\u6743\u91cd\u4e0b\u8f7d\/\u7f13\u5b58\u76ee\u5f55<\/td>\n<td>&#034;&#034;&#xff08;\u9ed8\u8ba4\u8def\u5f84&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>INPUT_LEN<\/td>\n<td>\u8bf7\u6c42\u8f93\u5165\u957f\u5ea6<\/td>\n<td>4000<\/td>\n<\/tr>\n<tr>\n<td>OUTPUT_LEN<\/td>\n<td>\u8bf7\u6c42\u8f93\u51fa\u957f\u5ea6<\/td>\n<td>16<\/td>\n<\/tr>\n<tr>\n<td>MAX_MODEL_LEN<\/td>\n<td>\u6a21\u578b\u6700\u5927\u957f\u5ea6<\/td>\n<td>4096<\/td>\n<\/tr>\n<tr>\n<td>MIN_CACHE_HIT_PCT<\/td>\n<td>\u524d\u7f00\u7f13\u5b58\u547d\u4e2d\u7387\u8981\u6c42&#xff0c;0\u2013100&#xff1b;\u8bbe\u4e3a 0 \u7981\u7528<\/td>\n<td>60<\/td>\n<\/tr>\n<tr>\n<td>MAX_LATENCY_ALLOWED_MS<\/td>\n<td>\u5141\u8bb8\u7684 P99 \u7aef\u5230\u7aef\u5ef6\u8fdf&#xff08;ms&#xff09;&#xff1b;\u8bbe\u6781\u5927\u503c\u53ef\u5ffd\u7565<\/td>\n<td>500<\/td>\n<\/tr>\n<tr>\n<td>NUM_SEQS_LIST<\/td>\n<td>\u5f85\u6d4b max-num-seqs \u5217\u8868<\/td>\n<td>&#034;128 256&#034;<\/td>\n<\/tr>\n<tr>\n<td>NUM_BATCHED_TOKENS_LIST<\/td>\n<td>\u5f85\u6d4b max-num-batched-tokens \u5217\u8868<\/td>\n<td>&#034;1024 2048 4096&#034;<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u77ed\u4e0a\u4e0b\u6587\u573a\u666f\u53ef\u9002\u5f53\u589e\u5927 max-num-seqs \u503c\u3002<\/p>\n<h4>1.3 \u8fd0\u884c\u6b65\u9aa4<\/h4>\n<li>\u914d\u7f6e&#xff1a;\u6309\u4e0a\u8868\u5728\u811a\u672c\u9876\u90e8\u4fee\u6539\u53d8\u91cf\u3002<\/li>\n<li>\u6267\u884c&#xff1a;<span class=\"token builtin class-name\">cd<\/span> <span class=\"token operator\">&lt;<\/span>\u811a\u672c\u6240\u5728\u76ee\u5f55<span class=\"token operator\">&gt;<\/span><br \/>\n<span class=\"token function\">bash<\/span> auto_tune.sh<\/p>\n<\/li>\n<p>\u6ce8\u610f&#xff1a;\u6267\u884c\u8def\u5f84\u4e2d\u4e0d\u80fd\u5305\u542b\u5b57\u7b26\u4e32 vllm&#xff0c;\u5426\u5219 pkill -f vllm \u4f1a\u8bef\u6740\u811a\u672c\u81ea\u8eab\u3002<\/p>\n<hr \/>\n<h3>2. \u8981\u70b9\u63d0\u70bc<\/h3>\n<h4>2.1 \u6838\u5fc3\u76ee\u6807<\/h4>\n<ul>\n<li>\u81ea\u52a8\u904d\u5386 max-num-seqs \u4e0e max-num-batched-tokens \u7ec4\u5408\u3002<\/li>\n<li>\u5728\u6ee1\u8db3\u5ef6\u8fdf\u6216\u7f13\u5b58\u547d\u4e2d\u7387\u7ea6\u675f\u7684\u524d\u63d0\u4e0b&#xff0c;\u627e\u5230\u6700\u5927\u541e\u5410\u3002<\/li>\n<\/ul>\n<h4>2.2 \u5178\u578b\u573a\u666f<\/h4>\n<table>\n<tr>\u76ee\u6807\u5173\u952e\u914d\u7f6e\u793a\u4f8b<\/tr>\n<tbody>\n<tr>\n<td>\u4ec5\u6700\u5927\u5316\u541e\u5410<\/td>\n<td>MAX_LATENCY_ALLOWED_MS&#061;1e11, MIN_CACHE_HIT_PCT&#061;0<\/td>\n<\/tr>\n<tr>\n<td>\u541e\u5410 &#043; \u5ef6\u8fdf\u7ea6\u675f<\/td>\n<td>MAX_LATENCY_ALLOWED_MS&#061;500<\/td>\n<\/tr>\n<tr>\n<td>\u541e\u5410 &#043; \u5ef6\u8fdf &#043; \u524d\u7f00\u7f13\u5b58<\/td>\n<td>MAX_LATENCY_ALLOWED_MS&#061;500, MIN_CACHE_HIT_PCT&#061;60<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>2.3 \u8f93\u51fa\u7ed3\u679c<\/h4>\n<ul>\n<li>\u4f4d\u4e8e $BASE\/auto-benchmark\/YYYY_MM_DD_HH_MM\/&#xff1a;\n<ul>\n<li>vllm_log_*.txt&#xff1a;\u5404\u53c2\u6570\u7ec4\u5408\u7684 vLLM \u65e5\u5fd7<\/li>\n<li>bm_log_*.txt&#xff1a;\u5bf9\u5e94 benchmark \u65e5\u5fd7<\/li>\n<li>result.txt&#xff1a;\u6700\u4f18\u53c2\u6570\u53ca\u541e\u5410\u6c47\u603b<\/li>\n<li>profile\/&#xff1a;\u6700\u4f73\u8fd0\u884c\u7684\u4e00\u6b21 profiler trace&#xff08;TPU \u4e3a .xplane.pb&#xff0c;GPU \u4e3a .json&#xff09;<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<hr \/>\n<h3>3. \u5982\u4f55\u8c03\u4f18 vLLM \u8fd0\u884c\u53c2\u6570&#xff08;\u5b9e\u6218\u6307\u5357&#xff09;<\/h3>\n<h4>3.1 \u8c03\u4f18\u6d41\u7a0b&#xff08;\u811a\u672c\u5185\u90e8\u903b\u8f91&#xff09;<\/h4>\n<li>\u786e\u5b9a\u6700\u5927 GPU \u5185\u5b58\u5229\u7528\u7387<br \/>\n\u4ece 0.98 \u5f00\u59cb\u9012\u51cf&#xff0c;\u9632\u6b62 OOM\u3002<\/li>\n<li>\u53cc\u91cd\u5faa\u73af\u904d\u5386<br \/>\n\u904d\u5386\u6240\u6709 (max-num-seqs, max-num-batched-tokens) \u7ec4\u5408\u3002<\/li>\n<li>\u5ef6\u8fdf\u611f\u77e5\u541e\u5410\u641c\u7d22\n<ul>\n<li>\u5148\u4ee5\u65e0\u9650\u8bf7\u6c42\u901f\u7387\u8dd1\u4e00\u8f6e&#xff1b;\u82e5 P99 \u5ef6\u8fdf\u6ee1\u8db3&#xff0c;\u5219\u8bb0\u5f55\u541e\u5410\u3002<\/li>\n<li>\u82e5\u5ef6\u8fdf\u8d85\u9650&#xff0c;\u5219\u9010\u6b65\u964d\u4f4e\u8bf7\u6c42\u901f\u7387&#xff0c;\u627e\u5230\u6ee1\u8db3\u5ef6\u8fdf\u7684\u6700\u9ad8\u541e\u5410\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u8bb0\u5f55\u6700\u4f18\u503c<br \/>\n\u6bcf\u6b21\u66f4\u65b0\u541e\u5410\u66f4\u9ad8\u7684\u6709\u6548\u7ec4\u5408\u3002<\/li>\n<li>\u4fdd\u5b58\u6027\u80fd\u753b\u50cf<br \/>\n\u5bf9\u6700\u4f73\u7ec4\u5408\u4fdd\u5b58 profiler trace&#xff0c;\u4fbf\u4e8e TensorBoard \u7b49\u5de5\u5177\u6df1\u5ea6\u5206\u6790\u3002<\/li>\n<h4>3.2 \u624b\u52a8\u5fae\u8c03\u5efa\u8bae<\/h4>\n<ul>\n<li>\u957f\u8f93\u5165 \/ \u957f\u8f93\u51fa\u573a\u666f\n<ul>\n<li>\u9002\u5f53\u964d\u4f4e max-num-seqs&#xff0c;\u63d0\u9ad8 max-num-batched-tokens&#xff0c;\u51cf\u5c11 padding \u6d6a\u8d39\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u77ed\u8f93\u5165 \/ \u77ed\u8f93\u51fa\u573a\u666f\n<ul>\n<li>\u63d0\u9ad8 max-num-seqs&#xff0c;\u964d\u4f4e max-num-batched-tokens&#xff0c;\u5145\u5206\u5229\u7528\u5e76\u53d1\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u663e\u5b58\u7d27\u5f20\n<ul>\n<li>\u964d\u4f4e gpu-memory-utilization \u6216 max-model-len\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u5ef6\u8fdf\u654f\u611f\n<ul>\n<li>\u5728 MAX_LATENCY_ALLOWED_MS \u8303\u56f4\u5185&#xff0c;\u4f18\u5148\u9009\u62e9\u541e\u5410\u6700\u9ad8\u7684\u7ec4\u5408&#xff0c;\u82e5\u4ecd\u8d85\u9650&#xff0c;\u5219\u964d\u4f4e max-num-seqs \u6216 max-num-batched-tokens\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u524d\u7f00\u7f13\u5b58\u4f18\u5316\n<ul>\n<li>\u82e5\u4e1a\u52a1\u6709\u5927\u91cf\u5171\u4eab\u524d\u7f00&#xff0c;\u53ef\u8bbe\u7f6e MIN_CACHE_HIT_PCT&gt;0&#xff0c;\u811a\u672c\u4f1a\u8fc7\u6ee4\u6389\u547d\u4e2d\u7387\u4e0d\u8fbe\u6807\u7684\u7ed3\u679c\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<p>\u811a\u672c\u5df2\u81ea\u52a8\u5316\u4e0a\u8ff0\u8fc7\u7a0b&#xff1b;\u5982\u60f3\u624b\u52a8\u5b9e\u9a8c&#xff0c;\u53ef\u76f4\u63a5\u7528 vllm serve \u542f\u52a8\u5e76\u914d\u5408 vllm bench serve \u8fdb\u884c\u57fa\u51c6\u6d4b\u8bd5\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb696\u6b21\uff0c\u70b9\u8d5e24\u6b21\uff0c\u6536\u85cf20\u6b21\u3002\u811a\u672c\u5df2\u81ea\u52a8\u5316\u4e0a\u8ff0\u8fc7\u7a0b\uff1b\u5982\u60f3\u624b\u52a8\u5b9e\u9a8c\uff0c\u53ef\u76f4\u63a5\u7528\u3002\u77ed\u4e0a\u4e0b\u6587\u573a\u666f\u53ef\u9002\u5f53\u589e\u5927\u3002_vllm tp \u542f\u52a8\u53c2\u6570<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[43,44],"topic":[],"class_list":["post-55836","post","type-post","status-publish","format-standard","hentry","category-server","tag-43","tag-44"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/55836.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb696\u6b21\uff0c\u70b9\u8d5e24\u6b21\uff0c\u6536\u85cf20\u6b21\u3002\u811a\u672c\u5df2\u81ea\u52a8\u5316\u4e0a\u8ff0\u8fc7\u7a0b\uff1b\u5982\u60f3\u624b\u52a8\u5b9e\u9a8c\uff0c\u53ef\u76f4\u63a5\u7528\u3002\u77ed\u4e0a\u4e0b\u6587\u573a\u666f\u53ef\u9002\u5f53\u589e\u5927\u3002_vllm tp \u542f\u52a8\u53c2\u6570\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/55836.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2025-08-14T00:54:44+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"1 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/55836.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/55836.html\",\"name\":\"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2025-08-14T00:54:44+00:00\",\"dateModified\":\"2025-08-14T00:54:44+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/55836.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/55836.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/55836.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/55836.html","og_locale":"zh_CN","og_type":"article","og_title":"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb696\u6b21\uff0c\u70b9\u8d5e24\u6b21\uff0c\u6536\u85cf20\u6b21\u3002\u811a\u672c\u5df2\u81ea\u52a8\u5316\u4e0a\u8ff0\u8fc7\u7a0b\uff1b\u5982\u60f3\u624b\u52a8\u5b9e\u9a8c\uff0c\u53ef\u76f4\u63a5\u7528\u3002\u77ed\u4e0a\u4e0b\u6587\u573a\u666f\u53ef\u9002\u5f53\u589e\u5927\u3002_vllm tp \u542f\u52a8\u53c2\u6570","og_url":"https:\/\/www.wsisp.com\/helps\/55836.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2025-08-14T00:54:44+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"1 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/55836.html","url":"https:\/\/www.wsisp.com\/helps\/55836.html","name":"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2025-08-14T00:54:44+00:00","dateModified":"2025-08-14T00:54:44+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/55836.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/55836.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/55836.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"\u81ea\u52a8\u8c03\u4f18 vLLM \u670d\u52a1\u5668\u53c2\u6570\uff08\u5b9e\u6218\u6307\u5357\uff09"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/55836","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=55836"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/55836\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=55836"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=55836"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=55836"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=55836"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}