{"id":34092,"date":"2025-07-28T23:41:27","date_gmt":"2025-07-28T15:41:27","guid":{"rendered":"https:\/\/www.kdjingpai.com\/?p=34092"},"modified":"2025-07-28T23:43:41","modified_gmt":"2025-07-28T15:43:41","slug":"glm-45","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/pt\/glm-45\/","title":{"rendered":"GLM-4.5\uff1a\u5f00\u6e90\u591a\u6a21\u6001\u5927\u6a21\u578b\u652f\u6301\u667a\u80fd\u63a8\u7406\u4e0e\u4ee3\u7801\u751f\u6210"},"content":{"rendered":"<p>GLM-4.5 \u662f zai-org \u5f00\u53d1\u7684\u4e00\u6b3e\u5f00\u6e90\u591a\u6a21\u6001\u5927\u8bed\u8a00\u6a21\u578b\uff0c\u4e13\u4e3a\u667a\u80fd\u63a8\u7406\u3001\u4ee3\u7801\u751f\u6210\u548c\u667a\u80fd\u4f53\u4efb\u52a1\u8bbe\u8ba1\u3002\u5b83\u5305\u542b GLM-4.5\uff083550 \u4ebf\u53c2\u6570\uff0c320 \u4ebf\u6d3b\u8dc3\u53c2\u6570\uff09\u3001GLM-4.5-Air\uff081060 \u4ebf\u53c2\u6570\uff0c120 \u4ebf\u6d3b\u8dc3\u53c2\u6570\uff09\u7b49\u591a\u4e2a\u53d8\u4f53\uff0c\u91c7\u7528\u6df7\u5408\u4e13\u5bb6\uff08MoE\uff09\u67b6\u6784\uff0c\u652f\u6301 128K \u4e0a\u4e0b\u6587\u957f\u5ea6\u548c 96K \u8f93\u51fa\u4ee4\u724c\u3002\u6a21\u578b\u5728 15 \u4e07\u4ebf\u4ee4\u724c\u4e0a\u9884\u8bad\u7ec3\uff0c\u7ecf\u8fc7\u4ee3\u7801\u3001\u63a8\u7406\u548c\u667a\u80fd\u4f53\u9886\u57df\u7684\u5fae\u8c03\uff0c\u6027\u80fd\u5728\u591a\u4e2a\u57fa\u51c6\u6d4b\u8bd5\u4e2d\u540d\u5217\u524d\u8305\uff0c\u7279\u522b\u662f\u5728\u7f16\u7a0b\u548c\u5de5\u5177\u8c03\u7528\u4efb\u52a1\u4e2d\u63a5\u8fd1\u751a\u81f3\u8d85\u8d8a\u90e8\u5206\u95ed\u6e90\u6a21\u578b\u3002GLM-4.5 \u4ee5 MIT \u8bb8\u53ef\u8bc1\u53d1\u5e03\uff0c\u652f\u6301\u5b66\u672f\u548c\u5546\u4e1a\u7528\u9014\uff0c\u9002\u5408\u5f00\u53d1\u8005\u3001\u7814\u7a76\u4eba\u5458\u548c\u4f01\u4e1a\u5728\u672c\u5730\u6216\u4e91\u7aef\u90e8\u7f72\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-34093\" src=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/07\/8112a0595303d8c-scaled.png\" alt=\"\" width=\"2560\" height=\"1758\" srcset=\"https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/07\/8112a0595303d8c-scaled.png 2560w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/07\/8112a0595303d8c-1536x1055.png 1536w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/07\/8112a0595303d8c-2048x1407.png 2048w, https:\/\/www.kdjingpai.com\/wp-content\/uploads\/2025\/07\/8112a0595303d8c-18x12.png 18w\" sizes=\"auto, (max-width: 2560px) 100vw, 2560px\" \/><\/p>\n<p>&nbsp;<\/p>\n<h2>\u529f\u80fd\u5217\u8868<\/h2>\n<ul>\n<li>\u6df7\u5408\u63a8\u7406\u6a21\u5f0f\uff1a\u652f\u6301\u201c\u601d\u8003\u6a21\u5f0f\u201d\u5904\u7406\u590d\u6742\u63a8\u7406\u548c\u5de5\u5177\u8c03\u7528\uff0c\u201c\u975e\u601d\u8003\u6a21\u5f0f\u201d\u63d0\u4f9b\u5feb\u901f\u54cd\u5e94\u3002<\/li>\n<li>\u591a\u6a21\u6001\u652f\u6301\uff1a\u5904\u7406\u6587\u672c\u548c\u56fe\u50cf\u8f93\u5165\uff0c\u9002\u7528\u4e8e\u591a\u6a21\u6001\u95ee\u7b54\u548c\u5185\u5bb9\u751f\u6210\u3002<\/li>\n<li>\u667a\u80fd\u7f16\u7a0b\uff1a\u751f\u6210 Python\u3001JavaScript \u7b49\u8bed\u8a00\u7684\u9ad8\u8d28\u91cf\u4ee3\u7801\uff0c\u652f\u6301\u4ee3\u7801\u8865\u5168\u548c bug \u4fee\u590d\u3002<\/li>\n<li>\u667a\u80fd\u4f53\u529f\u80fd\uff1a\u652f\u6301\u51fd\u6570\u8c03\u7528\u3001\u7f51\u9875\u6d4f\u89c8\u548c\u81ea\u52a8\u5316\u4efb\u52a1\u5904\u7406\uff0c\u9002\u5408\u590d\u6742\u5de5\u4f5c\u6d41\u3002<\/li>\n<li>\u4e0a\u4e0b\u6587\u7f13\u5b58\uff1a\u4f18\u5316\u957f\u5bf9\u8bdd\u6027\u80fd\uff0c\u51cf\u5c11\u91cd\u590d\u8ba1\u7b97\u3002<\/li>\n<li>\u7ed3\u6784\u5316\u8f93\u51fa\uff1a\u652f\u6301 JSON \u7b49\u683c\u5f0f\uff0c\u4fbf\u4e8e\u7cfb\u7edf\u96c6\u6210\u3002<\/li>\n<li>\u957f\u4e0a\u4e0b\u6587\u5904\u7406\uff1a\u539f\u751f\u652f\u6301 128K \u4e0a\u4e0b\u6587\u957f\u5ea6\uff0c\u9002\u5408\u957f\u6587\u6863\u5206\u6790\u3002<\/li>\n<li>\u6d41\u5f0f\u8f93\u51fa\uff1a\u63d0\u4f9b\u5b9e\u65f6\u54cd\u5e94\uff0c\u63d0\u5347\u4ea4\u4e92\u4f53\u9a8c\u3002<\/li>\n<\/ul>\n<h2>\u4f7f\u7528\u5e2e\u52a9<\/h2>\n<p>GLM-4.5 \u901a\u8fc7 GitHub \u4ed3\u5e93\uff08https:\/\/github.com\/zai-org\/GLM-4.5\uff09\u63d0\u4f9b\u6a21\u578b\u6743\u91cd\u548c\u5de5\u5177\uff0c\u9002\u5408\u6709\u6280\u672f\u80cc\u666f\u7684\u7528\u6237\u5728\u672c\u5730\u6216\u4e91\u7aef\u90e8\u7f72\u3002\u4ee5\u4e0b\u662f\u8be6\u7ec6\u7684\u5b89\u88c5\u548c\u4f7f\u7528\u6307\u5357\uff0c\u5e2e\u52a9\u7528\u6237\u5feb\u901f\u4e0a\u624b\u3002<\/p>\n<h3>\u5b89\u88c5\u6d41\u7a0b<\/h3>\n<ol>\n<li><strong>\u73af\u5883\u51c6\u5907<\/strong><br \/>\n\u786e\u4fdd\u5b89\u88c5 Python 3.8 \u6216\u4ee5\u4e0a\u7248\u672c\u548c Git\u3002\u5efa\u8bae\u4f7f\u7528\u865a\u62df\u73af\u5883\uff1a<\/p>\n<pre><code>python -m venv glm_env\r\nsource glm_env\/bin\/activate  # Linux\/Mac\r\nglm_env\\Scripts\\activate     # Windows\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u514b\u9686\u4ed3\u5e93<\/strong><br \/>\n\u4ece GitHub \u83b7\u53d6 GLM-4.5 \u4ee3\u7801\uff1a<\/p>\n<pre><code>git clone https:\/\/github.com\/zai-org\/GLM-4.5.git\r\ncd GLM-4.5\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u5b89\u88c5\u4f9d\u8d56<\/strong><br \/>\n\u5b89\u88c5\u6307\u5b9a\u7248\u672c\u7684\u4f9d\u8d56\u4ee5\u786e\u4fdd\u517c\u5bb9\u6027\uff1a<\/p>\n<pre><code>pip install setuptools&gt;=80.9.0 setuptools_scm&gt;=8.3.1\r\npip install git+https:\/\/github.com\/huggingface\/transformers.git@91221da2f1f68df9eb97c980a7206b14c4d3a9b0\r\npip install git+https:\/\/github.com\/vllm-project\/vllm.git@220aee902a291209f2975d4cd02dadcc6749ffe6\r\npip install torchvision&gt;=0.22.0 gradio&gt;=5.35.0 pre-commit&gt;=4.2.0 PyMuPDF&gt;=1.26.1 av&gt;=14.4.0 accelerate&gt;=1.6.0 spaces&gt;=0.37.1\r\n<\/code><\/pre>\n<p>\u6ce8\u610f\uff1avLLM \u7f16\u8bd1\u53ef\u80fd\u8017\u65f6\u8f83\u957f\uff0c\u82e5\u4e0d\u9700\u8981\u53ef\u4f7f\u7528\u9884\u7f16\u8bd1\u7248\u672c\u3002<\/li>\n<li><strong>\u6a21\u578b\u4e0b\u8f7d<\/strong><br \/>\n\u6a21\u578b\u6743\u91cd\u6258\u7ba1\u5728 Hugging Face \u548c ModelScope\u3002\u4ee5\u4e0b\u662f\u52a0\u8f7d GLM-4.5-Air \u7684\u793a\u4f8b\uff1a<\/p>\n<pre><code>from transformers import AutoTokenizer, AutoModel\r\ntokenizer = AutoTokenizer.from_pretrained(\"zai-org\/GLM-4.5-Air\", trust_remote_code=True)\r\nmodel = AutoModel.from_pretrained(\"zai-org\/GLM-4.5-Air\", trust_remote_code=True).half().cuda()\r\nmodel.eval()\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u786c\u4ef6\u8981\u6c42<\/strong>\n<ul>\n<li>GLM-4.5-Air\uff1a\u9700\u8981 16GB GPU \u5185\u5b58\uff08INT4 \u91cf\u5316\u7ea6 12GB\uff09\u3002<\/li>\n<li>GLM-4.5\uff1a\u5efa\u8bae\u591a GPU \u73af\u5883\uff0c\u7ea6\u9700 32GB \u5185\u5b58\u3002<\/li>\n<li>CPU \u63a8\u7406\uff1aGLM-4.5-Air \u53ef\u5728 32GB RAM \u7684 CPU \u4e0a\u8fd0\u884c\uff0c\u4f46\u901f\u5ea6\u8f83\u6162\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n<h3>\u4f7f\u7528\u65b9\u6cd5<\/h3>\n<p>GLM-4.5 \u652f\u6301\u547d\u4ee4\u884c\u3001\u7f51\u9875\u754c\u9762\u548c API \u8c03\u7528\uff0c\u63d0\u4f9b\u591a\u79cd\u4ea4\u4e92\u65b9\u5f0f\u3002<\/p>\n<h4>\u547d\u4ee4\u884c\u63a8\u7406<\/h4>\n<p>\u4f7f\u7528\u00a0<code>trans_infer_cli.py<\/code>\u00a0\u811a\u672c\u8fdb\u884c\u4ea4\u4e92\u5f0f\u5bf9\u8bdd\uff1a<\/p>\n<pre><code>python inference\/trans_infer_cli.py --model_name zai-org\/GLM-4.5-Air\r\n<\/code><\/pre>\n<ul>\n<li>\u8f93\u5165\u6587\u672c\u6216\u56fe\u50cf\uff0c\u6a21\u578b\u8fd4\u56de\u54cd\u5e94\u3002<\/li>\n<li>\u652f\u6301\u591a\u8f6e\u5bf9\u8bdd\uff0c\u81ea\u52a8\u4fdd\u5b58\u5386\u53f2\u8bb0\u5f55\u3002<\/li>\n<li>\u793a\u4f8b\uff1a\u751f\u6210 Python \u51fd\u6570\uff1a\n<pre><code>response, history = model.chat(tokenizer, \"\u5199\u4e00\u4e2a Python \u51fd\u6570\u8ba1\u7b97\u4e09\u89d2\u5f62\u9762\u79ef\", history=[])\r\nprint(response)\r\n<\/code><\/pre>\n<p>\u8f93\u51fa\uff1a<\/p>\n<pre><code>def triangle_area(base, height):\r\nreturn 0.5 * base * height\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>\u7f51\u9875\u754c\u9762<\/h4>\n<p>\u901a\u8fc7 Gradio \u542f\u52a8\u7f51\u9875\u754c\u9762\uff0c\u652f\u6301\u591a\u6a21\u6001\u8f93\u5165\uff1a<\/p>\n<pre><code>python inference\/trans_infer_gradio.py --model_name zai-org\/GLM-4.5-Air\r\n<\/code><\/pre>\n<ul>\n<li>\u8bbf\u95ee\u672c\u5730\u5730\u5740\uff08\u901a\u5e38\u4e3a\u00a0<code>http:\/\/127.0.0.1:7860<\/code>\uff09\u3002<\/li>\n<li>\u8f93\u5165\u6587\u672c\u6216\u4e0a\u4f20\u56fe\u7247\u3001PDF\uff0c\u70b9\u51fb\u63d0\u4ea4\u83b7\u53d6\u54cd\u5e94\u3002<\/li>\n<li>\u7279\u8272\u529f\u80fd\uff1a\u4e0a\u4f20 PDF\uff0c\u6a21\u578b\u53ef\u89e3\u6790\u5e76\u56de\u7b54\u95ee\u9898\u3002<\/li>\n<\/ul>\n<h4>API \u670d\u52a1<\/h4>\n<p>GLM-4.5 \u652f\u6301 OpenAI \u517c\u5bb9\u7684 API\uff0c\u4f7f\u7528 <a href=\"https:\/\/www.kdjingpai.com\/pt\/vllm\/\">vLLM<\/a> \u90e8\u7f72\uff1a<\/p>\n<pre><code>vllm serve zai-org\/GLM-4.5-Air --limit-mm-per-prompt '{\"image\":32}'\r\n<\/code><\/pre>\n<ul>\n<li>\u793a\u4f8b\u8bf7\u6c42\uff1a\n<pre><code>import requests\r\npayload = {\r\n\"model\": \"GLM-4.5-Air\",\r\n\"messages\": [{\"role\": \"user\", \"content\": \"\u5206\u6790\u8fd9\u5f20\u56fe\u7247\"}],\r\n\"image\": \"path\/to\/image.jpg\"\r\n}\r\nresponse = requests.post(\"http:\/\/localhost:8000\/v1\/chat\/completions\", json=payload)\r\nprint(response.json())\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<h4>\u7279\u8272\u529f\u80fd\u64cd\u4f5c<\/h4>\n<ol>\n<li><strong>\u6df7\u5408\u63a8\u7406\u6a21\u5f0f<\/strong>\n<ul>\n<li><strong>\u601d\u8003\u6a21\u5f0f<\/strong>\u00a0\uff1a\u9002\u5408\u590d\u6742\u4efb\u52a1\uff0c\u5982\u6570\u5b66\u63a8\u7406\u6216\u5de5\u5177\u8c03\u7528\uff1a<\/li>\n<\/ul>\n<pre><code>model.chat(tokenizer, \"\u89e3\u51b3\u65b9\u7a0b\uff1a2x^2 - 8x + 6 = 0\", mode=\"thinking\")\r\n<\/code><\/pre>\n<p>\u6a21\u578b\u4f1a\u8f93\u51fa\u8be6\u7ec6\u89e3\u9898\u6b65\u9aa4\u3002<\/p>\n<ul>\n<li><strong>\u975e\u601d\u8003\u6a21\u5f0f<\/strong>\u00a0\uff1a\u9002\u5408\u5feb\u901f\u95ee\u7b54\uff1a<\/li>\n<\/ul>\n<pre><code>model.chat(tokenizer, \"\u7ffb\u8bd1\uff1aGood morning\", mode=\"non-thinking\")\r\n<\/code><\/pre>\n<\/li>\n<li><strong>\u591a\u6a21\u6001\u652f\u6301<\/strong>\n<ul>\n<li>\u5904\u7406\u6587\u672c\u548c\u56fe\u50cf\u8f93\u5165\u3002\u4f8b\u5982\uff0c\u4e0a\u4f20\u6570\u5b66\u9898\u76ee\u56fe\u7247\uff1a\n<pre><code>python inference\/trans_infer_gradio.py --input math_problem.jpg\r\n<\/code><\/pre>\n<\/li>\n<li>\u6ce8\u610f\uff1a\u6682\u4e0d\u652f\u6301\u540c\u65f6\u5904\u7406\u56fe\u50cf\u548c\u89c6\u9891\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u667a\u80fd\u7f16\u7a0b<\/strong>\n<ul>\n<li>\u751f\u6210\u4ee3\u7801\uff1a\u8f93\u5165\u4efb\u52a1\u63cf\u8ff0\uff0c\u751f\u6210\u5b8c\u6574\u4ee3\u7801\uff1a\n<pre><code>response, _ = model.chat(tokenizer, \"\u5199\u4e00\u4e2a Python \u811a\u672c\u5b9e\u73b0\u8d2a\u5403\u86c7\u6e38\u620f\", history=[])\r\n<\/code><\/pre>\n<\/li>\n<li>\u652f\u6301\u4ee3\u7801\u8865\u5168\u548c bug \u4fee\u590d\uff0c\u9002\u5408\u5feb\u901f\u539f\u578b\u5f00\u53d1\u3002<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u4e0a\u4e0b\u6587\u7f13\u5b58<\/strong>\n<ul>\n<li>\u4f18\u5316\u957f\u5bf9\u8bdd\u6027\u80fd\uff0c\u51cf\u5c11\u91cd\u590d\u8ba1\u7b97\uff1a\n<pre><code>model.chat(tokenizer, \"\u7ee7\u7eed\u4e0a\u4e00\u8f6e\u5bf9\u8bdd\", cache_context=True)\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<\/li>\n<li><strong>\u7ed3\u6784\u5316\u8f93\u51fa<\/strong>\n<ul>\n<li>\u8f93\u51fa JSON \u683c\u5f0f\uff0c\u4fbf\u4e8e\u7cfb\u7edf\u96c6\u6210\uff1a\n<pre><code>response = model.chat(tokenizer, \"\u5217\u51fa Python \u7684\u57fa\u672c\u6570\u636e\u7c7b\u578b\", format=\"json\")\r\n<\/code><\/pre>\n<\/li>\n<\/ul>\n<\/li>\n<\/ol>\n<h3>\u6ce8\u610f\u4e8b\u9879<\/h3>\n<ul>\n<li>\u4f7f\u7528 transformers 4.49.0 \u53ef\u80fd\u6709\u517c\u5bb9\u6027\u95ee\u9898\uff0c\u63a8\u8350 4.48.3\u3002<\/li>\n<li>vLLM API \u5355\u6b21\u8f93\u5165\u6700\u591a\u652f\u6301 300 \u5f20\u56fe\u7247\u3002<\/li>\n<li>\u786e\u4fdd GPU \u9a71\u52a8\u652f\u6301 CUDA 11.8 \u6216\u4ee5\u4e0a\u3002<\/li>\n<\/ul>\n<h2>\u5e94\u7528\u573a\u666f<\/h2>\n<ol>\n<li><strong>\u7f51\u9875\u5f00\u53d1<\/strong><br \/>\nGLM-4.5 \u751f\u6210\u524d\u7aef\u548c\u540e\u7aef\u4ee3\u7801\uff0c\u652f\u6301\u5feb\u901f\u6784\u5efa\u73b0\u4ee3 Web \u5e94\u7528\u3002\u4f8b\u5982\uff0c\u521b\u5efa\u4ea4\u4e92\u5f0f\u7f51\u9875\u53ea\u9700\u51e0\u53e5\u63cf\u8ff0\u3002<\/li>\n<li><strong>\u667a\u80fd\u95ee\u7b54<\/strong><br \/>\n\u6a21\u578b\u89e3\u6790\u590d\u6742\u67e5\u8be2\uff0c\u7ed3\u5408\u7f51\u9875\u641c\u7d22\u548c\u77e5\u8bc6\u5e93\uff0c\u63d0\u4f9b\u7cbe\u51c6\u7b54\u6848\uff0c\u9002\u5408\u5ba2\u670d\u548c\u6559\u80b2\u573a\u666f\u3002<\/li>\n<li><strong>\u667a\u6167\u529e\u516c<\/strong><br \/>\n\u81ea\u52a8\u751f\u6210\u903b\u8f91\u6e05\u6670\u7684 PPT \u6216\u6d77\u62a5\uff0c\u652f\u6301\u4ece\u6807\u9898\u6269\u5c55\u5185\u5bb9\uff0c\u9002\u5408\u529e\u516c\u81ea\u52a8\u5316\u3002<\/li>\n<li><strong>\u4ee3\u7801\u751f\u6210<\/strong><br \/>\n\u751f\u6210 Python\u3001JavaScript \u7b49\u4ee3\u7801\uff0c\u652f\u6301\u591a\u8f6e\u8fed\u4ee3\u5f00\u53d1\uff0c\u9002\u5408\u5feb\u901f\u539f\u578b\u548c bug \u4fee\u590d\u3002<\/li>\n<li><strong>\u590d\u6742\u7ffb\u8bd1<\/strong><br \/>\n\u7ffb\u8bd1\u957f\u7bc7\u5b66\u672f\u6216\u653f\u7b56\u6587\u672c\uff0c\u4fdd\u6301\u8bed\u4e49\u4e00\u81f4\u6027\u548c\u98ce\u683c\uff0c\u9002\u5408\u51fa\u7248\u548c\u8de8\u5883\u670d\u52a1\u3002<\/li>\n<\/ol>\n<h2>QA<\/h2>\n<ol>\n<li><strong>GLM-4.5 \u548c GLM-4.5-Air \u7684\u533a\u522b\u662f\u4ec0\u4e48\uff1f<\/strong><br \/>\nGLM-4.5\uff083550 \u4ebf\u53c2\u6570\uff0c320 \u4ebf\u6d3b\u8dc3\uff09\u9002\u5408\u9ad8\u6027\u80fd\u63a8\u7406\uff1bGLM-4.5-Air\uff081060 \u4ebf\u53c2\u6570\uff0c120 \u4ebf\u6d3b\u8dc3\uff09\u66f4\u8f7b\u91cf\uff0c\u9002\u5408\u8d44\u6e90\u53d7\u9650\u73af\u5883\u3002<\/li>\n<li><strong>\u5982\u4f55\u4f18\u5316\u63a8\u7406\u901f\u5ea6\uff1f<\/strong><br \/>\n\u4f7f\u7528 GPU \u52a0\u901f\uff0c\u542f\u7528 INT4 \u91cf\u5316\uff0c\u6216\u9009\u62e9 GLM-4.5-Air \u964d\u4f4e\u8d44\u6e90\u9700\u6c42\u3002<\/li>\n<li><strong>\u662f\u5426\u652f\u6301\u5546\u4e1a\u7528\u9014\uff1f<\/strong><br \/>\n\u662f\u7684\uff0cMIT \u8bb8\u53ef\u8bc1\u5141\u8bb8\u514d\u8d39\u5546\u4e1a\u4f7f\u7528\u3002<\/li>\n<li><strong>\u5982\u4f55\u5904\u7406\u957f\u4e0a\u4e0b\u6587\uff1f<\/strong><br \/>\n\u539f\u751f\u652f\u6301 128K \u4e0a\u4e0b\u6587\uff0c\u542f\u7528\u00a0<code>yarn<\/code>\u00a0\u53c2\u6570\u53ef\u8fdb\u4e00\u6b65\u6269\u5c55\u3002<\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"<p>GLM-4.5 \u662f zai-org \u5f00\u53d1\u7684\u4e00\u6b3e\u5f00\u6e90\u591a\u6a21\u6001\u5927\u8bed\u8a00\u6a21\u578b\uff0c\u4e13\u4e3a\u667a\u80fd\u63a8\u7406\u3001\u4ee3\u7801\u751f\u6210\u548c\u667a\u80fd\u4f53\u4efb\u52a1\u8bbe\u8ba1\u3002\u5b83\u5305\u542b GLM-4.5\uff083550 \u4ebf\u53c2\u6570\uff0c320 \u4ebf\u6d3b\u8dc3\u53c2\u6570\uff09\u3001GLM-4.5-Air\uff081060 \u4ebf\u53c2\u6570\uff0c120 \u4ebf\u6d3b\u8dc3\u53c2\u6570\uff09\u7b49\u591a\u4e2a&#8230;<\/p>\n","protected":false},"author":1,"featured_media":32782,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20,392,397],"tags":[230],"class_list":["post-34092","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tool","category-models","category-text-model","tag-aikaiyuanxiangmu"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/posts\/34092","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/comments?post=34092"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/posts\/34092\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/media\/32782"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/media?parent=34092"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/categories?post=34092"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/pt\/wp-json\/wp\/v2\/tags?post=34092"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}