{"id":13,"date":"2024-07-19T12:57:08","date_gmt":"2024-07-19T04:57:08","guid":{"rendered":"https:\/\/www.onetts.com\/ai\/?p=13"},"modified":"2024-07-21T16:04:25","modified_gmt":"2024-07-21T08:04:25","slug":"qwen1-5-0-5b","status":"publish","type":"post","link":"https:\/\/www.onetts.com\/ai\/models\/qwen1-5-0-5b\/","title":{"rendered":"Qwen1.5-0.5B"},"content":{"rendered":"<p>Qwen1.5-0.5B \u662f\u4e00\u6b3e\u57fa\u4e8e Transformer \u67b6\u6784\u7684\u89e3\u7801\u5668\u6a21\u578b\uff0c\u7531 Qwen \u7cfb\u5217\u7684\u6700\u65b0\u6d4b\u8bd5\u7248 Qwen2 \u6f14\u5316\u800c\u6765\u3002\u8fd9\u6b3e\u6a21\u578b\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u9886\u57df\u5177\u6709\u663e\u8457\u7684\u6027\u80fd\u63d0\u5347\u548c\u591a\u8bed\u8a00\u652f\u6301\uff0c\u4f7f\u5176\u5728\u6587\u672c\u751f\u6210\u3001\u8bed\u8a00\u7406\u89e3\u548c\u804a\u5929\u673a\u5668\u4eba\u7b49\u9886\u57df\u5177\u6709\u5e7f\u6cdb\u7684\u5e94\u7528\u524d\u666f\u3002<\/p>\n<p>Qwen1.5 \u662f Qwen \u7cfb\u5217\u7684\u4e00\u6b21\u91cd\u5927\u5347\u7ea7\uff0c\u65e8\u5728\u63d0\u4f9b\u66f4\u9ad8\u6548\u3001\u66f4\u7a33\u5b9a\u7684\u8bed\u8a00\u6a21\u578b\u3002\u8be5\u7cfb\u5217\u6a21\u578b\u5305\u62ec\u591a\u79cd\u5927\u5c0f\uff0c\u4ece 0.5B\uff08\u5373 0.5 \u4ebf\u53c2\u6570\uff09\u5230 72B\uff08\u5373 720 \u4ebf\u53c2\u6570\uff09\uff0c\u6ee1\u8db3\u4e0d\u540c\u5e94\u7528\u573a\u666f\u7684\u9700\u6c42\u3002Qwen1.5-0.5B \u4f5c\u4e3a\u8be5\u7cfb\u5217\u4e2d\u7684\u8f83\u5c0f\u6a21\u578b\uff0c\u5c3d\u7ba1\u53c2\u6570\u91cf\u8f83\u5c11\uff0c\u4f46\u4f9d\u7136\u5728\u6027\u80fd\u4e0a\u8868\u73b0\u51fa\u8272\u3002<\/p>\n<h2>\u6a21\u578b\u8bc4\u6d4b<\/h2>\n<h3 id=\"-\">\u6280\u672f\u7279\u70b9<\/h3>\n<ol>\n<li><strong>\u591a\u6a21\u578b\u5c3a\u5bf8<\/strong>\uff1aQwen1.5 \u63d0\u4f9b\u4e86 8 \u79cd\u4e0d\u540c\u7684\u6a21\u578b\u5c3a\u5bf8\uff0c\u5305\u62ec 0.5B\u30011.8B\u30014B\u30017B\u300114B\u300132B \u548c 72B \u5bc6\u96c6\u6a21\u578b\uff0c\u4ee5\u53ca\u4e00\u4e2a 14B \u7684 MoE\uff08Mixture of Experts\uff09\u6a21\u578b\uff0c\u6fc0\u6d3b\u4e86 2.7B \u53c2\u6570\u3002\u8fd9\u79cd\u591a\u6837\u5316\u7684\u9009\u62e9\u4f7f\u5f97\u7528\u6237\u53ef\u4ee5\u6839\u636e\u5177\u4f53\u9700\u6c42\u9009\u62e9\u5408\u9002\u7684\u6a21\u578b\u5c3a\u5bf8\u3002<\/li>\n<li><strong>\u6027\u80fd\u63d0\u5347<\/strong>\uff1a\u4e0e\u4e4b\u524d\u7684 Qwen \u6a21\u578b\u76f8\u6bd4\uff0cQwen1.5 \u5728\u804a\u5929\u6a21\u578b\u7684\u6027\u80fd\u4e0a\u6709\u663e\u8457\u63d0\u5347\u3002\u8fd9\u4e3b\u8981\u5f97\u76ca\u4e8e\u5176\u4f18\u5316\u7684\u67b6\u6784\u548c\u8bad\u7ec3\u65b9\u6cd5\uff0c\u4f7f\u5176\u5728\u5904\u7406\u590d\u6742\u5bf9\u8bdd\u548c\u591a\u8f6e\u5bf9\u8bdd\u65f6\u66f4\u52a0\u6d41\u7545\u548c\u81ea\u7136\u3002<\/li>\n<li><strong>\u591a\u8bed\u8a00\u652f\u6301<\/strong>\uff1aQwen1.5 \u652f\u6301\u591a\u8bed\u8a00\uff0c\u5305\u62ec\u57fa\u7840\u8bed\u8a00\u6a21\u578b\u548c\u804a\u5929\u6a21\u578b\u3002\u8fd9\u4f7f\u5f97\u5b83\u80fd\u591f\u7406\u89e3\u548c\u751f\u6210\u591a\u79cd\u8bed\u8a00\u7684\u6587\u672c\uff0c\u6781\u5927\u5730\u6269\u5c55\u4e86\u5176\u5e94\u7528\u8303\u56f4\u3002<\/li>\n<li><strong>\u957f\u4e0a\u4e0b\u6587\u652f\u6301<\/strong>\uff1a\u6240\u6709 Qwen1.5 \u6a21\u578b\u5747\u652f\u6301 32K \u7684\u4e0a\u4e0b\u6587\u957f\u5ea6\uff0c\u8fd9\u610f\u5473\u7740\u5b83\u4eec\u80fd\u591f\u5904\u7406\u66f4\u957f\u7684\u6587\u672c\u8f93\u5165\uff0c\u4ece\u800c\u5728\u7406\u89e3\u548c\u751f\u6210\u6587\u672c\u65f6\u66f4\u52a0\u51c6\u786e\u548c\u8fde\u8d2f\u3002<\/li>\n<li><strong>\u65e0\u9700\u8fdc\u7a0b\u4ee3\u7801\u4fe1\u4efb<\/strong>\uff1aQwen1.5 \u4e0d\u9700\u8981\u00a0<code>trust_remote_code<\/code>\u00a0\u53c2\u6570\uff0c\u8fd9\u7b80\u5316\u4e86\u6a21\u578b\u7684\u90e8\u7f72\u548c\u4f7f\u7528\uff0c\u4f7f\u5f97\u7528\u6237\u53ef\u4ee5\u66f4\u8f7b\u677e\u5730\u5c06\u6a21\u578b\u96c6\u6210\u5230\u5404\u79cd\u5e94\u7528\u4e2d\u3002<\/li>\n<\/ol>\n<h3 id=\"-\">\u6a21\u578b\u67b6\u6784<\/h3>\n<p>Qwen1.5-0.5B \u57fa\u4e8e Transformer \u67b6\u6784\uff0c\u91c7\u7528\u4e86\u591a\u79cd\u5148\u8fdb\u7684\u6280\u672f\u6765\u63d0\u5347\u5176\u6027\u80fd\u3002\u4ee5\u4e0b\u662f\u5176\u4e3b\u8981\u67b6\u6784\u7279\u70b9\uff1a<\/p>\n<ul>\n<li><strong>SwiGLU \u6fc0\u6d3b<\/strong>\uff1aSwiGLU\uff08Switched Gated Linear Unit\uff09\u662f\u4e00\u79cd\u6fc0\u6d3b\u51fd\u6570\uff0c\u7528\u4e8e\u66ff\u4ee3\u4f20\u7edf\u7684 ReLU\uff08Rectified Linear Unit\uff09\u3002\u5b83\u5728\u5904\u7406\u590d\u6742\u6570\u636e\u65f6\u8868\u73b0\u66f4\u4f18\uff0c\u80fd\u591f\u66f4\u597d\u5730\u6355\u6349\u6570\u636e\u7684\u975e\u7ebf\u6027\u7279\u5f81\u3002<\/li>\n<li><strong>\u6ce8\u610f\u529b QKV \u504f\u7f6e<\/strong>\uff1a\u5728 Transformer \u6a21\u578b\u4e2d\uff0c\u6ce8\u610f\u529b\u673a\u5236\u662f\u6838\u5fc3\u3002Qwen1.5 \u901a\u8fc7\u5f15\u5165 QKV\uff08Query, Key, Value\uff09\u504f\u7f6e\uff0c\u4f18\u5316\u4e86\u6ce8\u610f\u529b\u673a\u5236\u7684\u8ba1\u7b97\u8fc7\u7a0b\uff0c\u63d0\u9ad8\u4e86\u6a21\u578b\u7684\u6548\u7387\u548c\u51c6\u786e\u6027\u3002<\/li>\n<li><strong>\u7ec4\u67e5\u8be2\u6ce8\u610f\u529b<\/strong>\uff1a\u8fd9\u79cd\u6280\u672f\u901a\u8fc7\u5c06\u67e5\u8be2\u5206\u7ec4\uff0c\u51cf\u5c11\u8ba1\u7b97\u91cf\uff0c\u540c\u65f6\u4fdd\u6301\u4e86\u6ce8\u610f\u529b\u673a\u5236\u7684\u51c6\u786e\u6027\u3002\u8fd9\u5bf9\u4e8e\u5904\u7406\u5927\u89c4\u6a21\u6570\u636e\u96c6\u5c24\u4e3a\u91cd\u8981\u3002<\/li>\n<li><strong>\u6ed1\u52a8\u7a97\u53e3\u4e0e\u5168\u6ce8\u610f\u529b\u6df7\u5408<\/strong>\uff1aQwen1.5 \u7ed3\u5408\u4e86\u6ed1\u52a8\u7a97\u53e3\u6ce8\u610f\u529b\u548c\u5168\u6ce8\u610f\u529b\uff0c\u4f7f\u5f97\u6a21\u578b\u5728\u5904\u7406\u957f\u6587\u672c\u65f6\u66f4\u52a0\u9ad8\u6548\uff0c\u540c\u65f6\u4fdd\u6301\u4e86\u5bf9\u5168\u5c40\u4e0a\u4e0b\u6587\u7684\u7406\u89e3\u80fd\u529b\u3002<\/li>\n<li><strong>\u6539\u8fdb\u7684\u5206\u8bcd\u5668<\/strong>\uff1aQwen1.5 \u914d\u5907\u4e86\u4e00\u4e2a\u6539\u8fdb\u7684\u5206\u8bcd\u5668\uff0c\u80fd\u591f\u9002\u5e94\u591a\u79cd\u81ea\u7136\u8bed\u8a00\u548c\u4ee3\u7801\uff0c\u63d0\u9ad8\u4e86\u6a21\u578b\u5728\u4e0d\u540c\u8bed\u8a00\u73af\u5883\u4e0b\u7684\u9002\u5e94\u6027\u3002<\/li>\n<\/ul>\n<h3 id=\"-\">\u5e94\u7528\u573a\u666f<\/h3>\n<p>Qwen1.5-0.5B \u7531\u4e8e\u5176\u51fa\u8272\u7684\u6027\u80fd\u548c\u7075\u6d3b\u6027\uff0c\u9002\u7528\u4e8e\u591a\u79cd\u81ea\u7136\u8bed\u8a00\u5904\u7406\u4efb\u52a1\u3002\u4ee5\u4e0b\u662f\u4e00\u4e9b\u5178\u578b\u7684\u5e94\u7528\u573a\u666f\uff1a<\/p>\n<ol>\n<li><strong>\u6587\u672c\u751f\u6210<\/strong>\uff1aQwen1.5 \u53ef\u4ee5\u7528\u4e8e\u751f\u6210\u9ad8\u8d28\u91cf\u7684\u6587\u672c\uff0c\u5982\u65b0\u95fb\u6587\u7ae0\u3001\u535a\u5ba2\u5e16\u5b50\u3001\u793e\u4ea4\u5a92\u4f53\u5185\u5bb9\u7b49\u3002\u5176\u591a\u8bed\u8a00\u652f\u6301\u4f7f\u5f97\u5b83\u80fd\u591f\u751f\u6210\u591a\u79cd\u8bed\u8a00\u7684\u6587\u672c\u3002<\/li>\n<li><strong>\u8bed\u8a00\u7406\u89e3<\/strong>\uff1aQwen1.5 \u5728\u7406\u89e3\u590d\u6742\u8bed\u8a00\u7ed3\u6784\u548c\u8bed\u5883\u65b9\u9762\u8868\u73b0\u51fa\u8272\uff0c\u53ef\u4ee5\u5e94\u7528\u4e8e\u673a\u5668\u7ffb\u8bd1\u3001\u60c5\u611f\u5206\u6790\u3001\u6587\u672c\u6458\u8981\u7b49\u4efb\u52a1\u3002<\/li>\n<li><strong>\u804a\u5929\u673a\u5668\u4eba<\/strong>\uff1aQwen1.5 \u7684\u804a\u5929\u6a21\u578b\u5728\u5904\u7406\u5bf9\u8bdd\u65f6\u66f4\u52a0\u81ea\u7136\u548c\u6d41\u7545\uff0c\u9002\u5408\u5f00\u53d1\u667a\u80fd\u5ba2\u670d\u3001\u865a\u62df\u52a9\u624b\u7b49\u804a\u5929\u673a\u5668\u4eba\u3002<\/li>\n<li><strong>\u5185\u5bb9\u63a8\u8350<\/strong>\uff1a\u5229\u7528 Qwen1.5 \u7684\u6587\u672c\u7406\u89e3\u80fd\u529b\uff0c\u53ef\u4ee5\u5f00\u53d1\u4e2a\u6027\u5316\u5185\u5bb9\u63a8\u8350\u7cfb\u7edf\uff0c\u4e3a\u7528\u6237\u63d0\u4f9b\u66f4\u7b26\u5408\u5176\u5174\u8da3\u548c\u504f\u597d\u7684\u5185\u5bb9\u3002<\/li>\n<li><strong>\u6559\u80b2\u8f85\u52a9<\/strong>\uff1aQwen1.5 \u53ef\u4ee5\u8f85\u52a9\u8bed\u8a00\u5b66\u4e60\uff0c\u63d0\u4f9b\u8bed\u6cd5\u7ea0\u6b63\u3001\u8bcd\u6c47\u63a8\u8350\u7b49\u529f\u80fd\uff0c\u5e2e\u52a9\u7528\u6237\u63d0\u9ad8\u8bed\u8a00\u80fd\u529b\u3002<\/li>\n<\/ol>\n<h2>\u90e8\u7f72\u4f7f\u7528<\/h2>\n<p>\u8981\u4f7f\u7528 Qwen1.5-0.5B \u6a21\u578b\uff0c\u9996\u5148\u9700\u8981\u5b89\u88c5 Hugging Face \u7684 transformers \u5e93\u3002\u5efa\u8bae\u5b89\u88c5\u7248\u672c\u4e3a\u00a0<code>transformers&gt;=4.37.0<\/code>\u3002\u4ee5\u4e0b\u662f\u5b89\u88c5\u6b65\u9aa4\uff1a<\/p>\n<ol>\n<li><strong>\u5b89\u88c5 transformers \u5e93<\/strong>\uff1a\n<pre><code class=\"lang-bash\">pip <span class=\"token function\">install<\/span> transformers\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u52a0\u8f7d Qwen1.5-0.5B \u6a21\u578b<\/strong>\uff1a\n<pre><code class=\"lang-python\"><span class=\"token keyword\">from<\/span> transformers <span class=\"token keyword\">import<\/span> AutoModel<span class=\"token punctuation\">,<\/span> AutoTokenizer\r\n\r\nmodel_name <span class=\"token operator\">=<\/span> <span class=\"token string\">\"qwen1.5-0.5b\"<\/span>\r\ntokenizer <span class=\"token operator\">=<\/span> AutoTokenizer<span class=\"token punctuation\">.<\/span>from_pretrained<span class=\"token punctuation\">(<\/span>model_name<span class=\"token punctuation\">)<\/span>\r\nmodel <span class=\"token operator\">=<\/span> AutoModel<span class=\"token punctuation\">.<\/span>from_pretrained<span class=\"token punctuation\">(<\/span>model_name<span class=\"token punctuation\">)<\/span>\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u6587\u672c\u751f\u6210\u793a\u4f8b<\/strong>\uff1a\n<pre><code class=\"lang-python\">input_text <span class=\"token operator\">=<\/span> <span class=\"token string\">\"The capital of France is\"<\/span>\r\noutput_text <span class=\"token operator\">=<\/span> model<span class=\"token punctuation\">.<\/span>generate<span class=\"token punctuation\">(<\/span>tokenizer<span class=\"token punctuation\">.<\/span>encode<span class=\"token punctuation\">(<\/span>input_text<span class=\"token punctuation\">,<\/span> return_tensors<span class=\"token operator\">=<\/span><span class=\"token string\">\"pt\"<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> max_length<span class=\"token operator\">=<\/span><span class=\"token number\">50<\/span><span class=\"token punctuation\">)<\/span>\r\n<span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span>tokenizer<span class=\"token punctuation\">.<\/span>decode<span class=\"token punctuation\">(<\/span>output_text<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span>\r\n<\/code><\/pre>\n<\/li>\n<\/ol>\n<h3 id=\"-\">\u5f15\u7528<\/h3>\n<p>\u5982\u679c\u4f60\u5728\u7814\u7a76\u6216\u5f00\u53d1\u4e2d\u4f7f\u7528\u4e86 Qwen1.5-0.5B \u6a21\u578b\uff0c\u5e76\u5e0c\u671b\u5f15\u7528\u8be5\u6a21\u578b\uff0c\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u5f15\u7528\u683c\u5f0f\uff1a<\/p>\n<pre><code>@article{qwen,\r\n  title={Qwen Technical Report},\r\n  author={Jinze Bai and Shuai Bai and Yunfei Chu and Zeyu Cui and Kai Dang and Xiaodong Deng and Yang Fan and Wenbin Ge and Yu Han and Fei Huang and Binyuan Hui and Luo Ji and Mei Li and Junyang Lin and Runji Lin and Dayiheng Liu and Gao Liu and Chengqiang Lu and Keming Lu and Jianxin Ma and Rui Men and Xingzhang Ren and Xuancheng Ren and Chuanqi Tan and Sinan Tan and Jianhong Tu and Peng Wang and Shijie Wang and Wei Wang and Shengguang Wu and Benfeng Xu and Jin Xu and An Yang and Hao Yang and Jian Yang and Shusheng Yang and Yang Yao and Bowen Yu and Hongyi Yuan and Zheng Yuan and Jianwei Zhang and Xingxuan Zhang and Yichang Zhang and Zhenru Zhang and Chang Zhou and Jingren Zhou and Xiaohuan Zhou and Tianhang Zhu},\r\n  journal={arXiv preprint arXiv:2309.16609},\r\n  year={2023}\r\n}\r\n<\/code><\/pre>\n<h2>\u5e38\u89c1\u95ee\u9898<\/h2>\n<p><strong>Q: Qwen1.5-0.5B \u6a21\u578b\u652f\u6301\u54ea\u4e9b\u8bed\u8a00\uff1f<\/strong><br \/>\nA: Qwen1.5-0.5B \u652f\u6301\u591a\u8bed\u8a00\uff0c\u5305\u62ec\u4f46\u4e0d\u9650\u4e8e\u82f1\u8bed\u3001\u4e2d\u6587\u3001\u897f\u73ed\u7259\u8bed\u3001\u6cd5\u8bed\u7b49\u3002\u5176\u591a\u8bed\u8a00\u652f\u6301\u4f7f\u5f97\u5b83\u80fd\u591f\u7406\u89e3\u548c\u751f\u6210\u591a\u79cd\u8bed\u8a00\u7684\u6587\u672c\u3002<\/p>\n<p><strong>Q: \u5982\u4f55\u5904\u7406\u957f\u6587\u672c\u8f93\u5165\uff1f<\/strong><br \/>\nA: Qwen1.5-0.5B \u652f\u6301 32K \u7684\u4e0a\u4e0b\u6587\u957f\u5ea6\uff0c\u53ef\u4ee5\u5904\u7406\u8f83\u957f\u7684\u6587\u672c\u8f93\u5165\u3002\u7528\u6237\u53ef\u4ee5\u901a\u8fc7\u8c03\u6574\u8f93\u5165\u6587\u672c\u7684\u957f\u5ea6\u6765\u5145\u5206\u5229\u7528\u8fd9\u4e00\u7279\u6027\u3002<\/p>\n<p><strong>Q: \u662f\u5426\u9700\u8981\u7279\u6b8a\u7684\u786c\u4ef6\u914d\u7f6e\uff1f<\/strong><br \/>\nA: Qwen1.5-0.5B \u53ef\u4ee5\u5728\u666e\u901a\u786c\u4ef6\u4e0a\u8fd0\u884c\uff0c\u4f46\u4e3a\u4e86\u83b7\u5f97\u66f4\u597d\u7684\u6027\u80fd\uff0c\u5efa\u8bae\u4f7f\u7528\u5177\u6709\u8f83\u9ad8\u8ba1\u7b97\u80fd\u529b\u7684\u786c\u4ef6\uff0c\u5982 GPU\u3002<\/p>\n<p><strong>Q: \u5982\u4f55\u5728\u804a\u5929\u673a\u5668\u4eba\u4e2d\u4f7f\u7528 Qwen1.5-0.5B\uff1f<\/strong><br \/>\nA: \u53ef\u4ee5\u901a\u8fc7\u52a0\u8f7d Qwen1.5-0.5B \u7684\u804a\u5929\u6a21\u578b\uff0c\u5e76\u5c06\u5176\u96c6\u6210\u5230\u804a\u5929\u673a\u5668\u4eba\u7684\u5bf9\u8bdd\u7ba1\u7406\u7cfb\u7edf\u4e2d\u3002\u6a21\u578b\u53ef\u4ee5\u5904\u7406\u7528\u6237\u7684\u8f93\u5165\u5e76\u751f\u6210\u76f8\u5e94\u7684\u56de\u590d\u3002<\/p>\n<p><strong>Q: \u5982\u4f55\u8fdb\u884c\u6a21\u578b\u7684\u5fae\u8c03\uff1f<\/strong><br \/>\nA: \u7528\u6237\u53ef\u4ee5\u6839\u636e\u81ea\u5df1\u7684\u9700\u6c42\u5bf9 Qwen1.5-0.5B \u8fdb\u884c\u5fae\u8c03\uff0c\u4f8b\u5982\u901a\u8fc7\u76d1\u7763\u5b66\u4e60\u3001\u5f3a\u5316\u5b66\u4e60\u6216\u7ee7\u7eed\u9884\u8bad\u7ec3\u7b49\u65b9\u6cd5\u3002<\/p>\n<h2>\u76f8\u5173\u8d44\u6e90\u94fe\u63a5<\/h2>\n<ul>\n<li><strong>Qwen1.5-0.5B \u6a21\u578b\u4e3b\u9875<\/strong>\uff1a<a target=\"_blank\" rel=\"noopener noreferrer nofollow\" href=\"https:\/\/modelscope.cn\/models\/qwen\/Qwen1.5-0.5B\" target=\"_blank\" rel=\"noopener\">Qwen1.5-0.5B<\/a><\/li>\n<li><strong>Hugging Face Transformers \u5e93<\/strong>\uff1a<a target=\"_blank\" rel=\"noopener noreferrer nofollow\" href=\"https:\/\/huggingface.co\/transformers\/\" target=\"_blank\" rel=\"noopener\">Hugging Face Transformers<\/a><\/li>\n<li><strong>Qwen \u6280\u672f\u62a5\u544a<\/strong>\uff1a<a target=\"_blank\" rel=\"noopener noreferrer nofollow\" href=\"https:\/\/arxiv.org\/abs\/2309.16609\" target=\"_blank\" rel=\"noopener\">Qwen Technical Report<\/a><\/li>\n<li><strong>GitHub \u4ed3\u5e93<\/strong>\uff1a<a target=\"_blank\" rel=\"noopener noreferrer nofollow\" href=\"https:\/\/github.com\/qwen-model\/qwen\" target=\"_blank\" rel=\"noopener\">Qwen GitHub<\/a><\/li>\n<\/ul>\n<h3 id=\"-\">\u7ed3\u8bba<\/h3>\n<p>Qwen1.5-0.5B \u6a21\u578b\u4f5c\u4e3a Qwen \u7cfb\u5217\u7684\u6700\u65b0\u6210\u5458\uff0c\u51ed\u501f\u5176\u5353\u8d8a\u7684\u6027\u80fd\u548c\u5e7f\u6cdb\u7684\u5e94\u7528\u524d\u666f\uff0c\u5df2\u7ecf\u6210\u4e3a\u81ea\u7136\u8bed\u8a00\u5904\u7406\u9886\u57df\u7684\u91cd\u8981\u5de5\u5177\u3002\u5176\u591a\u8bed\u8a00\u652f\u6301\u3001\u957f\u4e0a\u4e0b\u6587\u5904\u7406\u80fd\u529b\u548c\u4f18\u5316\u7684\u67b6\u6784\u8bbe\u8ba1\uff0c\u4f7f\u5176\u5728\u5404\u79cd\u8bed\u8a00\u4efb\u52a1\u4e2d\u8868\u73b0\u51fa\u8272\u3002\u5e0c\u671b\u672c\u6587\u7684\u4ecb\u7ecd\u80fd\u591f\u5e2e\u52a9\u4f60\u66f4\u597d\u5730\u7406\u89e3\u548c\u4f7f\u7528 Qwen1.5-0.5B \u6a21\u578b\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Qwen1.5-0.5B \u662f\u4e00\u6b3e\u57fa\u4e8e Transformer \u67b6\u6784\u7684\u89e3\u7801\u5668\u6a21\u578b\uff0c\u7531 Qwen \u7cfb\u5217\u7684\u6700\u65b0\u6d4b\u8bd5\u7248 Qwen2 \u6f14\u5316\u800c\u6765\u3002\u8fd9\u6b3e\u6a21\u578b\u5728\u81ea\u7136\u8bed\u8a00\u5904\u7406\u9886\u57df\u5177\u6709\u663e\u8457\u7684\u6027\u80fd\u63d0\u5347\u548c\u591a\u8bed\u8a00\u652f\u6301\uff0c\u4f7f\u5176\u5728\u6587\u672c\u751f\u6210\u3001\u8bed\u8a00\u7406\u89e3\u548c\u804a\u5929\u673a\u5668\u4eba\u7b49\u9886\u57df\u5177\u6709\u5e7f\u6cdb\u7684\u5e94\u7528\u524d\u666f\u3002 Qwen1.5 \u662f Qwen \u7cfb\u5217\u7684\u4e00\u6b21\u91cd\u5927\u5347\u7ea7\uff0c\u65e8\u5728\u63d0\u4f9b\u66f4\u9ad8\u6548\u3001\u66f4\u7a33\u5b9a\u7684\u8bed\u8a00\u6a21\u578b\u3002\u8be5\u7cfb\u5217\u6a21\u578b\u5305\u62ec\u591a\u79cd\u5927\u5c0f\uff0c\u4ece 0.5B\uff08\u5373 0.5 \u4ebf\u53c2\u6570\uff09\u5230 72B\uff08\u5373 720 \u4ebf\u53c2\u6570\uff09\uff0c\u6ee1\u8db3\u4e0d\u540c\u5e94\u7528\u573a\u666f\u7684\u9700\u6c42\u3002Qwen1.5-0.5B \u4f5c\u4e3a\u8be5\u7cfb\u5217\u4e2d\u7684\u8f83\u5c0f\u6a21\u578b\uff0c\u5c3d\u7ba1\u53c2\u6570\u91cf\u8f83\u5c11<\/p>\n","protected":false},"author":1,"featured_media":14,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"collection":[42],"company":[6],"rank":[],"class_list":["post-13","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-llm","collection-qwen1-5","company-alibaba"],"_links":{"self":[{"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/posts\/13","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/comments?post=13"}],"version-history":[{"count":4,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/posts\/13\/revisions"}],"predecessor-version":[{"id":31,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/posts\/13\/revisions\/31"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/media\/14"}],"wp:attachment":[{"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/media?parent=13"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/categories?post=13"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/tags?post=13"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/collection?post=13"},{"taxonomy":"company","embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/company?post=13"},{"taxonomy":"rank","embeddable":true,"href":"https:\/\/www.onetts.com\/ai\/wp-json\/wp\/v2\/rank?post=13"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}