{"id":14785,"date":"2024-12-04T18:52:07","date_gmt":"2024-12-04T10:52:07","guid":{"rendered":"https:\/\/www.aisharenet.com\/?p=14785"},"modified":"2024-12-04T18:52:07","modified_gmt":"2024-12-04T10:52:07","slug":"marco-o1","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/de\/marco-o1\/","title":{"rendered":"Marco-o1\uff1a\u57fa\u4e8eQwen2-7B-Instruct\u5fae\u8c03\u7684\u5f00\u6e90\u7248OpenAI o1\u6a21\u578b\uff0c\u63a2\u7d22\u5f00\u653e\u5f0f\u63a8\u7406\u6a21\u578b\uff0c\u89e3\u51b3\u590d\u6742\u95ee\u9898"},"content":{"rendered":"<p>Marco-o1\u662f\u7531\u963f\u91cc\u5df4\u5df4\u56fd\u9645\u6570\u5b57\u5546\u4e1a\u96c6\u56e2\uff08AIDC-AI\uff09\u5f00\u53d1\u7684\u5f00\u653e\u5f0f\u63a8\u7406\u6a21\u578b\uff0c\u65e8\u5728\u89e3\u51b3\u590d\u6742\u7684\u73b0\u5b9e\u4e16\u754c\u95ee\u9898\u3002\u8be5\u6a21\u578b\u7ed3\u5408\u4e86\u601d\u7ef4\u94fe\uff08CoT\uff09\u5fae\u8c03\u3001\u8499\u7279\u5361\u6d1b\u6811\u641c\u7d22\uff08MCTS\uff09\u548c\u521b\u65b0\u7684\u63a8\u7406\u7b56\u7565\uff0c\u4f18\u5316\u4e86\u590d\u6742\u95ee\u9898\u89e3\u51b3\u4efb\u52a1\u3002Marco-o1\u4e0d\u4ec5\u5173\u6ce8\u6570\u5b66\u3001\u7269\u7406\u548c\u7f16\u7a0b\u7b49\u6709\u6807\u51c6\u7b54\u6848\u7684\u5b66\u79d1\uff0c\u8fd8\u81f4\u529b\u4e8e\u5728\u6ca1\u6709\u660e\u786e\u6807\u51c6\u548c\u96be\u4ee5\u91cf\u5316\u5956\u52b1\u7684\u9886\u57df\u4e2d\u5b9e\u73b0\u6cdb\u5316\u3002\u8be5\u9879\u76ee\u7684\u76ee\u6807\u662f\u63a2\u7d22\u5927\u578b\u63a8\u7406\u6a21\u578b\u5728\u591a\u8bed\u8a00\u5e94\u7528\u4e2d\u7684\u6f5c\u529b\uff0c\u5e76\u901a\u8fc7\u4e0d\u65ad\u4f18\u5316\u548c\u6539\u8fdb\uff0c\u63d0\u5347\u6a21\u578b\u7684\u63a8\u7406\u80fd\u529b\u548c\u5e94\u7528\u8303\u56f4\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"aligncenter size-full wp-image-14786\" title=\"Marco-o1\uff1a\u57fa\u4e8eQwen2-7B-Instruct\u5fae\u8c03\u7684\u5f00\u6e90\u7248OpenAI o1\u6a21\u578b\uff0c\u63a2\u7d22\u5f00\u653e\u5f0f\u63a8\u7406\u6a21\u578b\uff0c\u89e3\u51b3\u590d\u6742\u95ee\u9898-1\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6.jpg\" alt=\"Marco-o1\uff1a\u57fa\u4e8eQwen2-7B-Instruct\u5fae\u8c03\u7684\u5f00\u6e90\u7248OpenAI o1\u6a21\u578b\uff0c\u63a2\u7d22\u5f00\u653e\u5f0f\u63a8\u7406\u6a21\u578b\uff0c\u89e3\u51b3\u590d\u6742\u95ee\u9898-1\" width=\"2256\" height=\"2038\" srcset=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6.jpg 2256w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-300x271.jpg 300w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-1024x925.jpg 1024w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-768x694.jpg 768w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-1536x1388.jpg 1536w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-2048x1850.jpg 2048w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2024\/12\/60cb1b551dcc2c6-13x12.jpg 13w\" sizes=\"auto, (max-width: 2256px) 100vw, 2256px\" \/><\/p>\n<p>&nbsp;<\/p>\n<h2>\u529f\u80fd\u5217\u8868<\/h2>\n<ul>\n<li><strong>\u601d\u7ef4\u94fe\u5fae\u8c03\uff08CoT\uff09<\/strong>\uff1a\u901a\u8fc7\u5bf9\u57fa\u7840\u6a21\u578b\u8fdb\u884c\u5168\u53c2\u6570\u5fae\u8c03\uff0c\u7ed3\u5408\u5f00\u6e90CoT\u6570\u636e\u96c6\u548c\u81ea\u7814\u5408\u6210\u6570\u636e\uff0c\u63d0\u5347\u6a21\u578b\u7684\u63a8\u7406\u80fd\u529b\u3002<\/li>\n<li><strong>\u8499\u7279\u5361\u6d1b\u6811\u641c\u7d22\uff08MCTS\uff09<\/strong>\uff1a\u5229\u7528\u6a21\u578b\u8f93\u51fa\u7684\u7f6e\u4fe1\u5ea6\u5f15\u5bfc\u641c\u7d22\uff0c\u6269\u5c55\u89e3\u51b3\u65b9\u6848\u7a7a\u95f4\uff0c\u4f18\u5316\u63a8\u7406\u8def\u5f84\u3002<\/li>\n<li><strong>\u63a8\u7406\u52a8\u4f5c\u7b56\u7565<\/strong>\uff1a\u5b9e\u65bd\u521b\u65b0\u7684\u63a8\u7406\u52a8\u4f5c\u7b56\u7565\u548c\u53cd\u601d\u673a\u5236\uff0c\u63a2\u7d22\u4e0d\u540c\u7c92\u5ea6\u7684\u52a8\u4f5c\uff0c\u63d0\u9ad8\u6a21\u578b\u89e3\u51b3\u590d\u6742\u95ee\u9898\u7684\u80fd\u529b\u3002<\/li>\n<li><strong>\u591a\u8bed\u8a00\u7ffb\u8bd1\u4efb\u52a1<\/strong>\uff1a\u9996\u6b21\u5c06\u5927\u578b\u63a8\u7406\u6a21\u578b\u5e94\u7528\u4e8e\u673a\u5668\u7ffb\u8bd1\u4efb\u52a1\uff0c\u63a2\u7d22\u591a\u8bed\u8a00\u548c\u7ffb\u8bd1\u9886\u57df\u7684\u63a8\u7406\u65f6\u95f4\u7f29\u653e\u89c4\u5f8b\u3002<\/li>\n<li><strong>\u5956\u52b1\u6a21\u578b\u8bad\u7ec3<\/strong>\uff1a\u5f00\u53d1\u7ed3\u679c\u5956\u52b1\u5efa\u6a21\uff08ORM\uff09\u548c\u8fc7\u7a0b\u5956\u52b1\u5efa\u6a21\uff08PRM\uff09\uff0c\u63d0\u4f9b\u66f4\u51c6\u786e\u7684\u5956\u52b1\u4fe1\u53f7\uff0c\u51cf\u5c11\u6811\u641c\u7d22\u7ed3\u679c\u7684\u968f\u673a\u6027\u3002<\/li>\n<li><strong>\u5f3a\u5316\u5b66\u4e60\u8bad\u7ec3<\/strong>\uff1a\u901a\u8fc7\u5f3a\u5316\u5b66\u4e60\u6280\u672f\uff0c\u4f18\u5316\u6a21\u578b\u7684\u51b3\u7b56\u8fc7\u7a0b\uff0c\u8fdb\u4e00\u6b65\u63d0\u5347\u5176\u95ee\u9898\u89e3\u51b3\u80fd\u529b\u3002<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<h2>\u4f7f\u7528\u5e2e\u52a9<\/h2>\n<h3>\u5b89\u88c5\u6d41\u7a0b<\/h3>\n<ol>\n<li><strong>\u8bbf\u95eeGitHub\u9875\u9762<\/strong>\uff1a\u524d\u5f80<a href=\"https:\/\/github.com\/AIDC-AI\/Marco-o1\">Marco-o1 GitHub\u9875\u9762<\/a>\u3002<\/li>\n<li><strong>\u514b\u9686\u4ed3\u5e93<\/strong>\uff1a\u4f7f\u7528\u547d\u4ee4<code>git clone https:\/\/github.com\/AIDC-AI\/Marco-o1.git<\/code>\u514b\u9686\u4ed3\u5e93\u5230\u672c\u5730\u3002<\/li>\n<li><strong>\u5b89\u88c5\u4f9d\u8d56<\/strong>\uff1a\u8fdb\u5165\u9879\u76ee\u76ee\u5f55\uff0c\u8fd0\u884c<code>pip install -r requirements.txt<\/code>\u5b89\u88c5\u6240\u9700\u4f9d\u8d56\u3002<\/li>\n<\/ol>\n<h3>\u4f7f\u7528\u6307\u5357<\/h3>\n<ol>\n<li><strong>\u52a0\u8f7d\u6a21\u578b<\/strong>\uff1a\u5728Python\u73af\u5883\u4e2d\uff0c\u4f7f\u7528\u4ee5\u4e0b\u4ee3\u7801\u52a0\u8f7d\u6a21\u578b\uff1a\n<pre><code>from transformers import AutoModelForCausalLM, AutoTokenizer\r\nmodel_name = \"AIDC-AI\/Marco-o1\"\r\nmodel = AutoModelForCausalLM.from_pretrained(model_name)\r\ntokenizer = AutoTokenizer.from_pretrained(model_name)\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u63a8\u7406\u793a\u4f8b<\/strong>\uff1a\u4f7f\u7528\u6a21\u578b\u8fdb\u884c\u63a8\u7406\uff0c\u4ee5\u4e0b\u662f\u4e00\u4e2a\u7b80\u5355\u7684\u793a\u4f8b\uff1a <code>python<br \/>\ninput_text = \"How many 'r' are in strawberry?\"<br \/>\ninputs = tokenizer(input_text, return_tensors=\"pt\")<br \/>\noutputs = model.generate(**inputs)<br \/>\nprint(tokenizer.decode(outputs[0], skip_special_tokens=True))<br \/>\n<\/code><\/li>\n<li><strong>\u591a\u8bed\u8a00\u7ffb\u8bd1<\/strong>\uff1aMarco-o1\u5728\u591a\u8bed\u8a00\u7ffb\u8bd1\u4efb\u52a1\u4e2d\u8868\u73b0\u51fa\u8272\uff0c\u4ee5\u4e0b\u662f\u4e00\u4e2a\u7ffb\u8bd1\u793a\u4f8b\uff1a <code>python<br \/>\ninput_text = \"\u8fd9\u4e2a\u978b\u62e5\u6709\u8e29\u5c4e\u611f\"<br \/>\ninputs = tokenizer(input_text, return_tensors=\"pt\")<br \/>\noutputs = model.generate(**inputs)<br \/>\nprint(tokenizer.decode(outputs[0], skip_special_tokens=True))<br \/>\n<\/code><\/li>\n<\/ol>\n<h3>\u8be6\u7ec6\u529f\u80fd\u64cd\u4f5c\u6d41\u7a0b<\/h3>\n<ol>\n<li><strong>\u601d\u7ef4\u94fe\u5fae\u8c03\uff08CoT\uff09<\/strong>\uff1a\n<ul>\n<li><strong>\u6570\u636e\u51c6\u5907<\/strong>\uff1a\u6536\u96c6\u5e76\u6574\u7406\u5f00\u6e90CoT\u6570\u636e\u96c6\u548c\u81ea\u7814\u5408\u6210\u6570\u636e\u3002<\/li>\n<li><strong>\u6a21\u578b\u5fae\u8c03<\/strong>\uff1a\u4f7f\u7528\u4e0a\u8ff0\u6570\u636e\u5bf9\u57fa\u7840\u6a21\u578b\u8fdb\u884c\u5168\u53c2\u6570\u5fae\u8c03\uff0c\u63d0\u5347\u5176\u63a8\u7406\u80fd\u529b\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u8499\u7279\u5361\u6d1b\u6811\u641c\u7d22\uff08MCTS\uff09<\/strong>\uff1a\n<ul>\n<li><strong>\u8282\u70b9\u8868\u793a<\/strong>\uff1a\u5728MCTS\u6846\u67b6\u4e2d\uff0c\u6bcf\u4e2a\u8282\u70b9\u8868\u793a\u95ee\u9898\u89e3\u51b3\u8fc7\u7a0b\u4e2d\u7684\u4e00\u4e2a\u63a8\u7406\u72b6\u6001\u3002<\/li>\n<li><strong>\u52a8\u4f5c\u8f93\u51fa<\/strong>\uff1a\u8282\u70b9\u7684\u53ef\u80fd\u52a8\u4f5c\u7531LLM\u751f\u6210\uff0c\u4ee3\u8868\u63a8\u7406\u94fe\u4e2d\u7684\u6f5c\u5728\u6b65\u9aa4\u3002<\/li>\n<li><strong>\u56de\u6eda\u548c\u5956\u52b1\u8ba1\u7b97<\/strong>\uff1a\u5728\u56de\u6eda\u9636\u6bb5\uff0cLLM\u7ee7\u7eed\u63a8\u7406\u8fc7\u7a0b\u76f4\u81f3\u7ec8\u6b62\u72b6\u6001\u3002<\/li>\n<li><strong>\u5f15\u5bfc\u641c\u7d22<\/strong>\uff1a\u4f7f\u7528\u5956\u52b1\u5206\u6570\u8bc4\u4f30\u548c\u9009\u62e9\u6709\u524d\u9014\u7684\u8def\u5f84\uff0c\u5f15\u5bfc\u641c\u7d22\u5411\u66f4\u53ef\u9760\u7684\u63a8\u7406\u94fe\u53d1\u5c55\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u63a8\u7406\u52a8\u4f5c\u7b56\u7565<\/strong>\uff1a\n<ul>\n<li><strong>\u52a8\u4f5c\u7c92\u5ea6<\/strong>\uff1a\u5728MCTS\u6846\u67b6\u5185\u63a2\u7d22\u4e0d\u540c\u7c92\u5ea6\u7684\u52a8\u4f5c\uff0c\u63d0\u9ad8\u641c\u7d22\u6548\u7387\u548c\u51c6\u786e\u6027\u3002<\/li>\n<li><strong>\u53cd\u601d\u673a\u5236<\/strong>\uff1a\u63d0\u793a\u6a21\u578b\u81ea\u6211\u53cd\u601d\uff0c\u663e\u8457\u589e\u5f3a\u5176\u89e3\u51b3\u590d\u6742\u95ee\u9898\u7684\u80fd\u529b\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u591a\u8bed\u8a00\u7ffb\u8bd1\u4efb\u52a1<\/strong>\uff1a\n<ul>\n<li><strong>\u4efb\u52a1\u5e94\u7528<\/strong>\uff1a\u5c06\u5927\u578b\u63a8\u7406\u6a21\u578b\u5e94\u7528\u4e8e\u673a\u5668\u7ffb\u8bd1\u4efb\u52a1\uff0c\u63a2\u7d22\u591a\u8bed\u8a00\u548c\u7ffb\u8bd1\u9886\u57df\u7684\u63a8\u7406\u65f6\u95f4\u7f29\u653e\u89c4\u5f8b\u3002<\/li>\n<li><strong>\u7ffb\u8bd1\u793a\u4f8b<\/strong>\uff1a\u5c55\u793a\u6a21\u578b\u5728\u7ffb\u8bd1\u4fda\u8bed\u8868\u8fbe\u65b9\u9762\u7684\u4f18\u8d8a\u8868\u73b0\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u5956\u52b1\u6a21\u578b\u8bad\u7ec3<\/strong>\uff1a\n<ul>\n<li><strong>\u7ed3\u679c\u5956\u52b1\u5efa\u6a21\uff08ORM\uff09<\/strong>\uff1a\u8bad\u7ec3\u6a21\u578b\u4ee5\u63d0\u4f9b\u66f4\u51c6\u786e\u7684\u5956\u52b1\u4fe1\u53f7\uff0c\u51cf\u5c11\u6811\u641c\u7d22\u7ed3\u679c\u7684\u968f\u673a\u6027\u3002<\/li>\n<li><strong>\u8fc7\u7a0b\u5956\u52b1\u5efa\u6a21\uff08PRM\uff09<\/strong>\uff1a\u901a\u8fc7\u8fc7\u7a0b\u5956\u52b1\u5efa\u6a21\uff0c\u8fdb\u4e00\u6b65\u4f18\u5316\u6a21\u578b\u7684\u63a8\u7406\u8def\u5f84\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u5f3a\u5316\u5b66\u4e60\u8bad\u7ec3<\/strong>\uff1a\n<ul>\n<li><strong>\u51b3\u7b56\u4f18\u5316<\/strong>\uff1a\u901a\u8fc7\u5f3a\u5316\u5b66\u4e60\u6280\u672f\uff0c\u4f18\u5316\u6a21\u578b\u7684\u51b3\u7b56\u8fc7\u7a0b\uff0c\u63d0\u5347\u5176\u95ee\u9898\u89e3\u51b3\u80fd\u529b\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"<p>Marco-o1\u662f\u7531\u963f\u91cc\u5df4\u5df4\u56fd\u9645\u6570\u5b57\u5546\u4e1a\u96c6\u56e2\uff08AIDC-AI\uff09\u5f00\u53d1\u7684\u5f00\u653e\u5f0f\u63a8\u7406\u6a21\u578b\uff0c\u65e8\u5728\u89e3\u51b3\u590d\u6742\u7684\u73b0\u5b9e\u4e16\u754c\u95ee\u9898\u3002\u8be5\u6a21\u578b\u7ed3\u5408\u4e86\u601d\u7ef4\u94fe\uff08CoT\uff09\u5fae\u8c03\u3001\u8499\u7279\u5361\u6d1b\u6811\u641c\u7d22\uff08MCTS\uff09\u548c\u521b\u65b0\u7684\u63a8\u7406\u7b56\u7565\uff0c\u4f18\u5316\u4e86\u590d\u6742\u95ee\u9898\u89e3\u51b3\u4efb\u52a1\u3002Marco-o1\u4e0d\u4ec5\u5173\u6ce8\u6570&#8230;<\/p>\n","protected":false},"author":1,"featured_media":61366,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[230],"class_list":["post-14785","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tool","tag-aikaiyuanxiangmu"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts\/14785","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/comments?post=14785"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/posts\/14785\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/media\/61366"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/media?parent=14785"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/categories?post=14785"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/de\/wp-json\/wp\/v2\/tags?post=14785"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}