{"id":131,"date":"2026-06-27T06:41:07","date_gmt":"2026-06-27T06:41:07","guid":{"rendered":"https:\/\/leetap.com\/blog\/2026\/06\/27\/qwen3-6-35b-a3b-%e8%b6%8a%e7%8b%b1%e7%89%88%e6%9d%a5%e4%ba%86%ef%bc%81%e6%9c%80%e5%bc%ba%e6%97%a0%e5%ae%a1%e6%9f%a5%e5%bc%80%e6%ba%90%e6%a8%a1%e5%9e%8b%ef%bc%8c6g%e6%98%be%e5%ad%98%e9%83%bd%e8%83%bd\/"},"modified":"2026-06-30T12:15:29","modified_gmt":"2026-06-30T12:15:29","slug":"qwen3-6-35b-a3b-%e8%b6%8a%e7%8b%b1%e7%89%88%e6%9d%a5%e4%ba%86%ef%bc%81%e6%9c%80%e5%bc%ba%e6%97%a0%e5%ae%a1%e6%9f%a5%e5%bc%80%e6%ba%90%e6%a8%a1%e5%9e%8b%ef%bc%8c6g%e6%98%be%e5%ad%98%e9%83%bd%e8%83%bd","status":"publish","type":"post","link":"https:\/\/leetap.com\/blog\/2026\/06\/27\/qwen3-6-35b-a3b-%e8%b6%8a%e7%8b%b1%e7%89%88%e6%9d%a5%e4%ba%86%ef%bc%81%e6%9c%80%e5%bc%ba%e6%97%a0%e5%ae%a1%e6%9f%a5%e5%bc%80%e6%ba%90%e6%a8%a1%e5%9e%8b%ef%bc%8c6g%e6%98%be%e5%ad%98%e9%83%bd%e8%83%bd\/","title":{"rendered":"Qwen3.6-35B-A3B \u8d8a\u72f1\u7248\u6765\u4e86\uff01\u6700\u5f3a\u65e0\u5ba1\u67e5\u5f00\u6e90\u6a21\u578b\uff0c6G\u663e\u5b58\u90fd\u80fd\u8dd1"},"content":{"rendered":"\r\n

Qwen3.6-35B-A3B \u8d8a\u72f1\u7248\u6765\u4e86\uff016G\u663e\u5b58\u5c31\u80fd\u8dd1\u7684”\u65e0\u5ba1\u67e5”\u6700\u5f3a\u5f00\u6e90\u6a21\u578b\uff0c\u9644\u5b8c\u6574\u90e8\u7f72\u6559\u7a0b<\/h1>\r\n\r\n\r\n

\u5144\u5f1f\u4eec\uff0cAI \u5708\u6700\u8fd1\u70b8\u4e86\u4e00\u4e2a\u6d88\u606f\u2014\u2014\u901a\u4e49\u5343\u95ee\u51fa\u4e86\u4e00\u4e2a”\u8d8a\u72f1\u7248”\u5f00\u6e90\u6a21\u578b Qwen3.6-35B-A3B<\/strong>\u3002<\/p>\r\n\r\n

\u8bf4\u5b83”\u8d8a\u72f1”\uff0c\u4e0d\u662f\u90a3\u79cd\u8fdd\u6cd5\u7ffb\u5899\u7684\u8d8a\u72f1\uff0c\u800c\u662f \u53bb\u6389\u4e86\u5b89\u5168\u8fc7\u6ee4\u5668<\/strong>\u3002\u666e\u901a\u6a21\u578b\u4f60\u95ee\u5b83”\u5e2e\u6211\u5199\u4e00\u6bb5\u53cd\u6d3e\u53d1\u8a00””\u5199\u4e00\u4e2a\u66b4\u529b\u7684\u6253\u6597\u573a\u666f”\uff0c\u5b83\u53ef\u80fd\u4f1a\u62d2\u7edd\u3002\u8fd9\u4e2a\u7248\u672c\u4e0d\u7ba1\u8fd9\u4e9b\uff0c\u56de\u7b54\u66f4\u653e\u5f97\u5f00\uff0c\u7f16\u7a0b\u3001\u5199\u4f5c\u3001\u521b\u610f\u7c7b\u4efb\u52a1\u4e0a\u5c24\u5176\u597d\u7528\u3002<\/p>\r\n\r\n

\u6700\u72e0\u7684\u662f\u4ec0\u4e48\u5462\uff1f\u5b83\u867d\u7136\u603b\u53c2\u6570 35B\uff0c\u4f46\u7528\u4e86 MoE\uff08\u6df7\u5408\u4e13\u5bb6\uff09\u67b6\u6784<\/strong>\u2014\u2014\u4f60\u53ef\u4ee5\u7406\u89e3\u4e3a\u4e00\u4e2a 35 \u4eba\u7684\u961f\u4f0d\uff0c\u6bcf\u6b21\u53ea\u53eb\u6700\u64c5\u957f\u7684 3 \u4e2a\u4eba\u5e72\u6d3b\u3002\u6240\u4ee5 6GB \u663e\u5b58\u5c31\u80fd\u8dd1\uff01<\/strong> \u6bd4\u4f60\u60f3\u8c61\u4e2d\u9700\u8981\u7684\u914d\u7f6e\u4f4e\u5f97\u591a\u3002<\/p>\r\n\r\n

\u800c\u4e14\u5b83\u662f\u591a\u6a21\u6001\u6a21\u578b\u2014\u2014\u56fe\u7247\u4e5f\u80fd\u770b\u3002\u4f60\u7529\u4e00\u5f20\u56fe\u8fc7\u53bb\uff0c\u5b83\u5e2e\u4f60\u5206\u6790\u3002<\/p>\r\n\r\n

\u4eca\u5929\u6559\u7a0b\u5c31\u628a\u8fd9\u4e2a”\u65e0\u5ba1\u67e5\u602a\u517d”\u8dd1\u8d77\u6765\u30026G \u663e\u5b58\u5c31\u591f\u4e86\u3002<\/p>\r\n\r\n\r\n

\u4e00\u3001\u5148\u770b\u4f60\u7684\u914d\u7f6e\u591f\u4e0d\u591f<\/h2>\r\n\r\n\r\n

\u8fd9\u4e2a\u6a21\u578b\u5bf9\u786c\u4ef6\u6781\u5176\u53cb\u597d\uff1a<\/p>\r\n\r\n\r\n

\r\n\r\n\r\n\r\n\r\n\r\n\r\n
\u786c\u4ef6<\/th>\r\n\u6700\u4f4e\u914d\u7f6e<\/th>\r\n\u63a8\u8350\u914d\u7f6e<\/th>\r\n<\/tr>\r\n<\/thead>\r\n
\u663e\u5361<\/strong><\/td>\r\n6GB \u663e\u5b58\uff08NVIDIA\uff09<\/td>\r\nRTX 3060 12G \/ RTX 4060<\/td>\r\n<\/tr>\r\n
\u7cfb\u7edf\u5185\u5b58<\/strong><\/td>\r\n8GB<\/td>\r\n16GB \u4ee5\u4e0a<\/td>\r\n<\/tr>\r\n
\u786c\u76d8<\/strong><\/td>\r\n10GB \u7a7a\u95f2<\/td>\r\nSSD\uff0c20GB+<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table><\/figure>\r\n\r\n\r\n

\u770b\u5230\u4e86\u5417\uff1f\u6700\u4f4e 6GB \u663e\u5b58\u5c31\u591f\u4e86\u3002<\/strong> \u4f60\u7684 GTX 1660 Super\uff086G\uff09\u3001RTX 2060\uff086G\uff09\u3001RTX 3060\uff0812G\uff09\u3001\u751a\u81f3\u7b14\u8bb0\u672c\u7684 RTX 3050 Ti\uff084G \u4e0d\u884c\uff0c\u5dee\u4e00\u70b9\uff09\u90fd\u80fd\u4e0a\u3002<\/p>\r\n\r\n

\ud83d\udc41\ufe0f \u67e5\u914d\u7f6e\uff1a<\/strong>\u6309 Win + R<\/code> \u2192 \u8f93 cmd<\/code> \u2192 \u5728\u9ed1\u6846\u91cc\u8f93 nvidia-smi<\/code> \u2192 \u770b\u6700\u4e0a\u9762\u90a3\u884c\uff0c\u786e\u8ba4\u663e\u5b58 \u2265 6GB\u3002<\/p>\r\n\r\n\r\n

\u4e8c\u3001\u65b9\u6848A\uff1aOllama \u90e8\u7f72\uff08\u65b0\u624b\u63a8\u8350\uff0c3\u5206\u949f\u4e0a\u624b\uff09<\/h2>\r\n\r\n\r\n

\u8fd9\u4e2a\u65b9\u6848\u6700\u7b80\u5355\uff0c\u4e0d\u7528\u5199 Python \u4ee3\u7801\u3002\u88c5\u4e00\u4e2a Ollama\uff0c\u4e00\u884c\u547d\u4ee4\u62c9\u6a21\u578b\u5c31\u80fd\u8dd1\u3002<\/p>\r\n\r\n

\u7b2c1\u6b65\uff1a\u5b89\u88c5 Ollama<\/h3>\r\n\r\n

\u505a\u4ec0\u4e48\uff1a<\/strong>Ollama \u662f\u5e2e\u4f60\u7ba1\u7406\u672c\u5730\u5927\u6a21\u578b\u7684\u5de5\u5177\u7bb1\u2014\u2014\u4e0b\u8f7d\u3001\u8fd0\u884c\u3001\u7ba1\u7406\u4e00\u6761\u9f99\u3002<\/p>\r\n\r\n

\u600e\u4e48\u505a\uff1a<\/strong><\/p>\r\n

    \r\n
  1. \u6253\u5f00\u6d4f\u89c8\u5668\uff0c\u8bbf\u95ee ollama.com<\/a><\/li>\r\n
  2. \u70b9 Download\uff0c\u4e0b Windows \u7248<\/li>\r\n
  3. \u53cc\u51fb\u5b89\u88c5 \u2192 \u4e00\u8def Next \u2192 Finish<\/li>\r\n<\/ol>\r\n\r\n

    \ud83d\udc41\ufe0f \u9a8c\u8bc1\uff1a<\/strong>\u6309 Win + R<\/code> \u2192 \u8f93 cmd<\/code> \u2192 \u8f93\u5165\uff1a<\/p>\r\n

    \r\nollama --version\r\n<\/pre>\r\n
    \u770b\u5230\u7248\u672c\u53f7\uff08\u6bd4\u5982 v0.5.4\uff09\u5c31\u5bf9\u4e86\u3002<\/p>\r\n\r\n
    \u7b2c2\u6b65\uff1a\u62c9\u53d6 Qwen3.6-35B-A3B \u6a21\u578b<\/h3>\r\n\r\n
    Ollama \u793e\u533a\u5df2\u7ecf\u6253\u5305\u597d\u4e86\u4f18\u5316\u7248\uff0c\u76f4\u63a5\u4e00\u884c\u547d\u4ee4\u4e0b\u8f7d\uff1a<\/p>\r\n\r\n
    \r\nollama pull qwen3.6:35b-a3b\r\n<\/pre>\r\n\r\n
    \ud83d\udc41\ufe0f \u4e0b\u8f7d\u8981\u591a\u4e45\uff1f<\/strong> \u6a21\u578b\u5927\u6982 6GB\uff08\u56e0\u4e3a MoE \u67b6\u6784\uff0c\u867d\u7136\u603b\u53c2\u6570 35B\uff0c\u4f46\u5b9e\u9645\u6587\u4ef6\u5f88\u5c0f\uff09\u3002\u4e00\u822c 5-15 \u5206\u949f\u4e0b\u5b8c\u3002<\/p>\r\n\r\n
    \r\n\u26a0\ufe0f \u6ce8\u610f\uff1a<\/strong>\u5982\u679c\u62a5 “manifest not found”\uff0c\u8bf4\u660e Ollama \u4ed3\u5e93\u91cc\u7684\u6807\u7b7e\u540d\u53ef\u80fd\u6709\u53d8\u5316\u3002\u53bb ollama.com\/library\/qwen3.6<\/a> \u641c\u4e00\u4e0b\u6700\u65b0\u6807\u7b7e\u540d\uff0c\u628a\u547d\u4ee4\u91cc\u7684 35b-a3b<\/code> \u6362\u6210\u6700\u65b0\u7684\u3002\r\n<\/p>\r\n\r\n
    \u7b2c3\u6b65\uff1a\u8dd1\u8d77\u6765<\/h3>\r\n\r\n
    \r\nollama run qwen3.6:35b-a3b\r\n<\/pre>\r\n\r\n
    \u7b49\u51e0\u79d2\u949f\u52a0\u8f7d\u5b8c\u6210\uff0c\u4f60\u5c31\u80fd\u76f4\u63a5\u6253\u5b57\u8ddf\u5b83\u804a\u5929\u4e86\u3002<\/p>\r\n\r\n
    \ud83d\udc41\ufe0f \u9a8c\u8bc1\uff1a<\/strong>\u8f93\u5165\u4e00\u4e2a\u6bd4\u8f83”\u653e\u5f97\u5f00”\u7684\u95ee\u9898\u8bd5\u8bd5\uff0c\u6bd4\u5982”\u5e2e\u6211\u5199\u4e00\u6bb5\u53cd\u6d3e\u89d2\u8272\u7684\u72ec\u767d”\u3002\u666e\u901a\u6a21\u578b\u53ef\u80fd\u4f1a\u62d2\u7edd\uff0c\u4f46 A3B \u7248\u672c\u4f1a\u76f4\u63a5\u7ed9\u4f60\u5199\u3002<\/p>\r\n\r\n
    \u591a\u6a21\u6001\uff08\u770b\u56fe\u529f\u80fd\uff09\uff1a<\/strong>\u8fd9\u6a21\u578b\u80fd\u770b\u56fe\u3002\u62d6\u4e00\u5f20\u56fe\u7247\u5230\u7ec8\u7aef\u7a97\u53e3\uff0c\u7136\u540e\u8f93\u5165\uff1a<\/p>\r\n
    \r\n>>> \/Users\/yourname\/Desktop\/chart.png  \u5206\u6790\u8fd9\u5f20\u56fe\u8868\u7684\u6570\u636e\r\n<\/pre>\r\n\r\n\r\n
    \u4e09\u3001\u65b9\u6848B\uff1aHuggingFace \u624b\u52a8\u90e8\u7f72\uff08\u8fdb\u9636\uff09<\/h2>\r\n\r\n\r\n
    \u5982\u679c\u4f60\u4e0d\u60f3\u7528 Ollama\uff0c\u60f3\u81ea\u5df1\u63a7\u5236\u91cf\u5316\u53c2\u6570\u3001\u6216\u8005\u4f60\u9700\u8981\u96c6\u6210\u5230\u81ea\u5df1\u7684 Python \u9879\u76ee\u91cc\uff0c\u7528\u8fd9\u4e2a\u65b9\u6848\u3002<\/p>\r\n\r\n
    \u7b2c1\u6b65\uff1a\u4e0b\u8f7d\u6a21\u578b\u6587\u4ef6<\/h3>\r\n
    \r\npip install huggingface-hub\r\n\r\nhuggingface-cli download Qwen\/Qwen3.6-35B-A3B \\\r\n  --local-dir .\/models\/qwen3.6-35b-a3b\r\n<\/pre>\r\n\r\n
    \u7b2c2\u6b65\uff1a4-bit \u91cf\u5316\u52a0\u8f7d\uff08\u5403\u663e\u5b58\u6700\u5c11\uff09<\/h3>\r\n
    \r\npip install transformers torch accelerate bitsandbytes\r\n\r\npython3 -c \"\r\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\r\nimport torch\r\n\r\nmodel_name = 'Qwen\/Qwen3.6-35B-A3B'\r\ntokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)\r\n\r\n# 4-bit \u91cf\u5316\uff0c6GB \u663e\u5b58\u4e5f\u80fd\u8dd1\r\nmodel = AutoModelForCausalLM.from_pretrained(\r\n    model_name, torch_dtype=torch.float16,\r\n    device_map='auto', load_in_4bit=True,\r\n    trust_remote_code=True\r\n)\r\n\r\nprompt = '\u7528100\u5b57\u89e3\u91ca\u4ec0\u4e48\u662f\u5927\u8bed\u8a00\u6a21\u578b'\r\ninputs = tokenizer(prompt, return_tensors='pt').to('cuda')\r\noutput = model.generate(**inputs, max_new_tokens=200)\r\nprint(tokenizer.decode(output[0], skip_special_tokens=True))\r\n\"\r\n<\/pre>\r\n\r\n\r\n
    \u56db\u3001A3B \u548c\u666e\u901a\u7248\u6709\u4ec0\u4e48\u533a\u522b\uff1f<\/h2>\r\n\r\n\r\n
    \u641e\u6df7\u4e86\uff1f\u522b\u6025\uff0c\u544a\u8bc9\u4f60\u533a\u522b\uff1a<\/p>\r\n\r\n\r\n
    \r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n
    \u7279\u6027<\/th>\r\n\u666e\u901a\u7248\uff08Qwen3.6-35B\uff09<\/th>\r\nA3B \u8d8a\u72f1\u7248<\/th>\r\n<\/tr>\r\n<\/thead>\r\n
    \u663e\u5b58\u8981\u6c42<\/strong><\/td>\r\n8GB\uff08\u91cf\u5316\u540e\uff09<\/td>\r\n6GB<\/td>\r\n<\/tr>\r\n
    \u67b6\u6784<\/strong><\/td>\r\nDense\uff08\u6240\u6709\u53c2\u6570\u90fd\u6fc0\u6d3b\uff09<\/td>\r\nMoE\uff08\u6bcf\u6b21\u53ea\u6fc0\u6d3b3B\uff09<\/td>\r\n<\/tr>\r\n
    \u5b89\u5168\u8fc7\u6ee4<\/strong><\/td>\r\n\u6709\uff0c\u654f\u611f\u5185\u5bb9\u62d2\u7edd\u56de\u7b54<\/td>\r\n\u65e0\uff0c\u56de\u7b54\u66f4\u653e\u5f97\u5f00<\/td>\r\n<\/tr>\r\n
    \u6a21\u578b\u6587\u4ef6\u5927\u5c0f<\/strong><\/td>\r\n~20GB\uff08\u91cf\u5316\u540e\uff09<\/td>\r\n~6GB<\/td>\r\n<\/tr>\r\n
    \u9002\u5408\u573a\u666f<\/strong><\/td>\r\n\u901a\u7528\u95ee\u7b54\u3001\u4f01\u4e1a\u5e94\u7528<\/td>\r\n\u521b\u610f\u5199\u4f5c\u3001\u7f16\u7a0b\u3001\u89d2\u8272\u626e\u6f14<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table><\/figure>\r\n\r\n\r\n
    \u7b80\u5355\u8bf4\uff1a\u4f60\u8981\u505a\u521b\u4f5c\u3001\u5199\u4ee3\u7801\u3001\u641e\u89d2\u8272\u626e\u6f14 \u2192 \u9009 A3B \u7248\u3002\u4f60\u8981\u505a\u4f01\u4e1a\u5e94\u7528\u3001\u9700\u8981\u5185\u5bb9\u5b89\u5168\u8fc7\u6ee4 \u2192 \u9009\u666e\u901a\u7248\u3002<\/strong><\/p>\r\n\r\n\r\n
    \u4e94\u3001\u5e38\u89c1\u95ee\u9898<\/h2>\r\n\r\n\r\n
    Q\uff1a\u663e\u5b58\u6ea2\u51fa\u4e86\u600e\u4e48\u529e\uff1f<\/h3>\r\n
    \u68c0\u67e5\u4e24\u70b9\uff1a\u2460 \u786e\u8ba4\u7528\u4e86 4-bit \u91cf\u5316\uff08load_in_4bit=True\uff09<\/code> \u2461 \u5173\u6389\u6d4f\u89c8\u5668\u3001\u6e38\u620f\u7b49\u5403\u663e\u5b58\u7684\u7a0b\u5e8f\u3002\u5982\u679c\u53ea\u6709 4GB \u663e\u5b58\uff0c\u6362 Qwen3.6-8B\uff088B \u5c0f\u6a21\u578b\uff09\u3002<\/p>\r\n\r\n
    Q\uff1a\u4e0b\u8f7d\u6a21\u578b\u592a\u6162\u4e86\u600e\u4e48\u529e\uff1f<\/h3>\r\n
    HuggingFace \u56fd\u5185\u4e0b\u8f7d\u6162\u662f\u901a\u75c5\u3002\u89e3\u51b3\u65b9\u6848\uff1a<\/p>\r\n
    \r\n# \u8bbe\u955c\u50cf\u7ad9\r\nexport HF_ENDPOINT=https:\/\/hf-mirror.com\r\n<\/pre>\r\n
    \u6216\u8005\u5728 Ollama \u91cc\u4e0b\u2014\u2014Ollama \u7684\u4e0b\u8f7d\u901f\u5ea6\u901a\u5e38\u6bd4 HuggingFace \u5feb\u3002<\/p>\r\n\r\n
    Q\uff1aOllama \u62a5\u9519 “manifest not found”<\/h3>\r\n
    \u53bb ollama.com\/library\/qwen3.6<\/a> \u770b\u6700\u65b0\u7684\u6807\u7b7e\u540d\uff0c\u66ff\u6362\u6389\u547d\u4ee4\u91cc\u7684\u6a21\u578b\u540d\u5c31\u884c\u3002\u6216\u8005\u6362\u65b9\u6848B\u7528 HuggingFace\u3002<\/p>\r\n\r\n
    Q\uff1aA3B \u7248\u771f\u7684”\u8d8a\u72f1”\u4e86\uff1f\u4f1a\u4e0d\u4f1a\u8fdd\u89c4\uff1f<\/h3>\r\n
    \u8fd9\u4e2a”\u8d8a\u72f1”\u6307\u7684\u662f\u53bb\u6389\u4e86\u5b89\u5168\u8fc7\u6ee4\u5668\uff0c\u4e0d\u662f\u7834\u89e3\u6216\u76d7\u7248\u3002\u5b83\u662f\u963f\u91cc\u901a\u4e49\u5343\u95ee\u5b98\u65b9\u53d1\u5e03\u7684\u6a21\u578b\uff0cApache 2.0 \u5f00\u6e90\u534f\u8bae\uff0c\u53ef\u4ee5\u514d\u8d39\u5546\u7528<\/strong>\u3002\u5b83\u4e0d\u8fdd\u6cd5\uff0c\u53ea\u662f\u56de\u7b54\u4e0d\u4f1a\u50cf\u666e\u901a AI \u90a3\u6837”\u6211\u662f\u4e00\u4e2a AI \u52a9\u624b\uff0c\u6211\u4e0d\u80fd\u2026\u2026”\u3002<\/p>\r\n\r\n
    Q\uff1a\u8fd9\u4e2a\u6a21\u578b\u548c GPT-5 \u6bd4\u600e\u4e48\u6837\uff1f<\/h3>\r\n
    \u5728\u6570\u5b66\u3001\u7f16\u7a0b\u65b9\u9762\u63a5\u8fd1 GPT-5 \u6c34\u5e73\uff0c\u521b\u610f\u5199\u4f5c\u5404\u6709\u5343\u79cb\u3002\u4f46\u5b83\u514d\u8d39\u3001\u5f00\u6e90\u3001\u672c\u5730\u8dd1<\/strong>\uff0c\u6ca1\u6709\u9690\u79c1\u6cc4\u9732\u98ce\u9669\u3002GPT-5 \u8981\u4ed8\u8d39\uff0c\u800c\u4e14\u4f60\u7684\u804a\u5929\u8bb0\u5f55\u4f1a\u4e0a\u4f20\u5230 OpenAI \u670d\u52a1\u5668\u3002\u770b\u4f60\u5728\u4e4e\u4ec0\u4e48\u4e86\u3002<\/p>\r\n\r\n\r\n
    \u516d\u3001\u603b\u7ed3<\/h2>\r\n\r\n\r\n
    Qwen3.6-35B-A3B \u662f 6GB \u663e\u5b58\u80fd\u8dd1\u7684\u6700\u5f3a\u5f00\u6e90\u6a21\u578b\u4e4b\u4e00\u3002<\/strong>\u6ca1\u6709\u4e4b\u4e00\u3002<\/p>\r\n
      \r\n
    1. \u65b0\u624b\u8d70 Ollama<\/strong>\uff1a\u4e00\u884c\u547d\u4ee4 ollama pull qwen3.6:35b-a3b<\/code>\uff0c3 \u5206\u949f\u4e0a\u624b<\/li>\r\n
    2. \u8fdb\u9636\u8d70 HuggingFace<\/strong>\uff1a\u81ea\u5df1\u63a7\u5236\u91cf\u5316\u53c2\u6570\uff0c\u7cbe\u5ea6\u66f4\u9ad8<\/li>\r\n
    3. \u591a\u6a21\u6001\u662f\u4eae\u70b9<\/strong>\uff1a\u80fd\u770b\u56fe\u5206\u6790\u56fe\u8868\u548c\u56fe\u7247<\/li>\r\n
    4. \u65e0\u5ba1\u67e5<\/strong>\uff1a\u521b\u610f\u5199\u4f5c\u548c\u7f16\u7a0b\u573a\u666f\u975e\u5e38\u723d<\/li>\r\n<\/ol>\r\n\r\n
      \r\n\ud83d\udca1 \u7ed9\u65b0\u624b\u7684\u8bdd\uff1a<\/strong>\u522b\u7ea0\u7ed3\u65b9\u6848A\u8fd8\u662f\u65b9\u6848B\u3002\u88c5 Ollama\uff0c\u4e00\u884c\u547d\u4ee4\u62c9\u6a21\u578b\uff0c\u7528\u4e86\u518d\u8bf4\u3002\u8fd9\u6a21\u578b\u5c0f\u5230 6G \u5361\u90fd\u80fd\u8dd1\uff0c\u4f60\u663e\u5361\u522b\u518d\u5403\u7070\u4e86\uff01<\/p>\r\n","protected":false},"excerpt":{"rendered":"
      Qwen3.6-35B-A3B \u8d8a\u72f1\u7248\u6765\u4e86\uff016G\u663e\u5b58\u5c31\u80fd\u8dd1\u7684”\u65e0\u5ba1\u67e5”\u6700\u5f3a\u5f00\u6e90\u6a21\u578b\uff0c\u9644… <\/p>\n","protected":false},"author":0,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"neve_meta_sidebar":"","neve_meta_container":"","neve_meta_enable_content_width":"","neve_meta_content_width":0,"neve_meta_title_alignment":"","neve_meta_author_avatar":"","neve_post_elements_order":"","neve_meta_disable_header":"","neve_meta_disable_footer":"","neve_meta_disable_title":"","footnotes":""},"categories":[39],"tags":[],"class_list":["post-131","post","type-post","status-publish","format-standard","hentry","category-qwen"],"_links":{"self":[{"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/posts\/131","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/comments?post=131"}],"version-history":[{"count":4,"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/posts\/131\/revisions"}],"predecessor-version":[{"id":154,"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/posts\/131\/revisions\/154"}],"wp:attachment":[{"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/media?parent=131"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/categories?post=131"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/leetap.com\/blog\/wp-json\/wp\/v2\/tags?post=131"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}