{"id":27608,"date":"2025-03-04T15:40:58","date_gmt":"2025-03-04T07:40:58","guid":{"rendered":"https:\/\/www.aisharenet.com\/?p=27608"},"modified":"2025-03-04T15:40:58","modified_gmt":"2025-03-04T07:40:58","slug":"cogview4","status":"publish","type":"post","link":"https:\/\/www.kdjingpai.com\/en\/cogview4\/","title":{"rendered":"CogView4\uff1a\u751f\u6210\u4e2d\u82f1\u53cc\u8bed\u9ad8\u6e05\u56fe\u7247\u7684\u5f00\u6e90\u6587\u751f\u56fe\u6a21\u578b"},"content":{"rendered":"

CogView4 \u662f\u7531\u6e05\u534e\u5927\u5b66 KEG \u5b9e\u9a8c\u5ba4\uff08THUDM\uff09\u5f00\u53d1\u7684\u4e00\u6b3e\u5f00\u6e90\u6587\u751f\u56fe\u6a21\u578b\uff0c\u4e13\u6ce8\u4e8e\u5c06\u6587\u672c\u63cf\u8ff0\u8f6c\u5316\u4e3a\u9ad8\u8d28\u91cf\u56fe\u50cf\u3002\u5b83\u652f\u6301\u4e2d\u82f1\u53cc\u8bed\u63d0\u793a\u8bcd\u8f93\u5165\uff0c\u5c24\u5176\u64c5\u957f\u7406\u89e3\u4e2d\u6587\u63d0\u793a\u5e76\u751f\u6210\u5e26\u6709\u6c49\u5b57\u7684\u56fe\u50cf\uff0c\u975e\u5e38\u9002\u5408\u5e7f\u544a\u8bbe\u8ba1\u3001\u77ed\u89c6\u9891\u521b\u4f5c\u7b49\u573a\u666f\u3002\u4f5c\u4e3a\u9996\u4e2a\u652f\u6301\u5728\u753b\u9762\u4e2d\u751f\u6210\u6c49\u5b57\u7684\u5f00\u6e90\u6a21\u578b\uff0cCogView4 \u5728\u590d\u6742\u8bed\u4e49\u5bf9\u9f50\u548c\u6307\u4ee4\u8ddf\u968f\u80fd\u529b\u4e0a\u8868\u73b0\u51fa\u8272\u3002\u5b83\u57fa\u4e8e GLM-4-9B \u6587\u672c\u7f16\u7801\u5668\uff0c\u652f\u6301\u4efb\u610f\u957f\u5ea6\u7684\u63d0\u793a\u8bcd\u8f93\u5165\uff0c\u5e76\u80fd\u751f\u6210\u9ad8\u8fbe 2048 \u5206\u8fa8\u7387\u7684\u56fe\u50cf\u3002\u9879\u76ee\u6258\u7ba1\u5728 GitHub \u4e0a\uff0c\u63d0\u4f9b\u8be6\u7ec6\u4ee3\u7801\u548c\u4f7f\u7528\u6587\u6863\uff0c\u5438\u5f15\u4e86\u5927\u91cf\u5f00\u53d1\u8005\u4e0e\u521b\u4f5c\u8005\u5173\u6ce8\u4e0e\u53c2\u4e0e\u3002<\/p>\n

\u6700\u65b0\u7684 CogView4 \u6a21\u578b\u5c06\u4e8e3\u670813\u65e5\u4e0a\u7ebf \u667a\u8c31\u6e05\u8a00<\/a> \u5b98\u7f51\u3002<\/p><\/blockquote>\n

\"CogView4\uff1a\u751f\u6210\u4e2d\u82f1\u53cc\u8bed\u9ad8\u6e05\u56fe\u7247\u7684\u5f00\u6e90\u6587\u751f\u56fe\u6a21\u578b-1\"

\u5728\u7ebf\u4f53\u9a8c\uff1ahttps:\/\/huggingface.co\/spaces\/THUDM-HF-SPACE\/CogView4<\/p><\/div>\n

 <\/p>\n

\u529f\u80fd\u5217\u8868<\/h2>\n
    \n
  • \u4e2d\u82f1\u53cc\u8bed\u63d0\u793a\u8bcd\u751f\u6210\u56fe\u50cf<\/strong>\uff1a\u652f\u6301\u4e2d\u6587\u548c\u82f1\u6587\u63cf\u8ff0\uff0c\u80fd\u7cbe\u51c6\u7406\u89e3\u5e76\u751f\u6210\u7b26\u5408\u63d0\u793a\u7684\u56fe\u50cf\uff0c\u4e2d\u6587\u573a\u666f\u8868\u73b0\u5c24\u4e3a\u7a81\u51fa\u3002<\/li>\n
  • \u753b\u9762\u751f\u6210\u6c49\u5b57<\/strong>\uff1a\u53ef\u5728\u56fe\u50cf\u4e2d\u751f\u6210\u6e05\u6670\u7684\u4e2d\u6587\u6587\u5b57\uff0c\u9002\u5408\u5236\u4f5c\u6d77\u62a5\u3001\u5e7f\u544a\u7b49\u9700\u8981\u6587\u5b57\u5185\u5bb9\u7684\u521b\u610f\u4f5c\u54c1\u3002<\/li>\n
  • \u4efb\u610f\u5206\u8fa8\u7387\u8f93\u51fa<\/strong>\uff1a\u652f\u6301\u751f\u6210\u4ece\u4f4e\u5206\u8fa8\u7387\u5230 2048×2048 \u7684\u4efb\u610f\u5c3a\u5bf8\u56fe\u50cf\uff0c\u6ee1\u8db3\u591a\u6837\u5316\u9700\u6c42\u3002<\/li>\n
  • \u8d85\u957f\u63d0\u793a\u8bcd\u652f\u6301<\/strong>\uff1a\u63a5\u53d7\u4efb\u610f\u957f\u5ea6\u7684\u6587\u672c\u8f93\u5165\uff0c\u6700\u591a\u53ef\u5904\u7406 1024 \u4e2a token\uff0c\u4fbf\u4e8e\u63cf\u8ff0\u590d\u6742\u573a\u666f\u3002<\/li>\n
  • \u590d\u6742\u8bed\u4e49\u5bf9\u9f50<\/strong>\uff1a\u80fd\u51c6\u786e\u6355\u6349\u63d0\u793a\u8bcd\u4e2d\u7684\u7ec6\u8282\uff0c\u751f\u6210\u7b26\u5408\u8bed\u4e49\u7684\u9ad8\u8d28\u91cf\u56fe\u50cf\u3002<\/li>\n
  • \u5f00\u6e90\u6a21\u578b\u5b9a\u5236<\/strong>\uff1a\u63d0\u4f9b\u5b8c\u6574\u4ee3\u7801\u548c\u9884\u8bad\u7ec3\u6a21\u578b\uff0c\u5f00\u53d1\u8005\u53ef\u6839\u636e\u9700\u6c42\u8fdb\u884c\u4e8c\u6b21\u5f00\u53d1\u6216\u4f18\u5316\u3002<\/li>\n<\/ul>\n

     <\/p>\n

    \u4f7f\u7528\u5e2e\u52a9<\/h2>\n

    \u5b89\u88c5\u6d41\u7a0b<\/h3>\n

    CogView4 \u662f\u4e00\u4e2a\u57fa\u4e8e Python \u7684\u5f00\u6e90\u9879\u76ee\uff0c\u9700\u8981\u5728\u672c\u5730\u914d\u7f6e\u73af\u5883\u624d\u80fd\u8fd0\u884c\u3002\u4ee5\u4e0b\u662f\u8be6\u7ec6\u7684\u5b89\u88c5\u6b65\u9aa4\uff1a<\/p>\n

    1. \u73af\u5883\u51c6\u5907<\/h4>\n
      \n
    • \u64cd\u4f5c\u7cfb\u7edf<\/strong>\uff1a\u652f\u6301 Windows\u3001Linux \u6216 macOS\u3002<\/li>\n
    • \u786c\u4ef6\u8981\u6c42<\/strong>\uff1a\u5efa\u8bae\u914d\u5907 NVIDIA GPU\uff08\u81f3\u5c11 16GB \u663e\u5b58\uff09\u4ee5\u52a0\u901f\u63a8\u7406\uff0cCPU \u4e5f\u53ef\u8fd0\u884c\u4f46\u901f\u5ea6\u8f83\u6162\u3002<\/li>\n
    • \u8f6f\u4ef6\u4f9d\u8d56<\/strong>\uff1a\n
        \n
      • Python 3.8 \u6216\u66f4\u9ad8\u7248\u672c<\/li>\n
      • PyTorch\uff08\u63a8\u8350\u5b89\u88c5 GPU \u7248\u672c\uff0ctorch>=2.0\uff09<\/li>\n
      • Git\uff08\u7528\u4e8e\u514b\u9686\u4ed3\u5e93\uff09<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n

        2. \u514b\u9686\u4ed3\u5e93<\/h4>\n

        \u6253\u5f00\u7ec8\u7aef\uff0c\u8f93\u5165\u4ee5\u4e0b\u547d\u4ee4\u4e0b\u8f7d CogView4 \u9879\u76ee\u6e90\u7801\uff1a<\/p>\n

        git clone https:\/\/github.com\/THUDM\/CogView4.git  \r\ncd CogView4\r\n<\/code><\/pre>\n
        3. \u5b89\u88c5\u4f9d\u8d56<\/h4>\n
        \u9879\u76ee\u63d0\u4f9b requirements.txt \u6587\u4ef6\uff0c\u8fd0\u884c\u4ee5\u4e0b\u547d\u4ee4\u5b89\u88c5\u6240\u9700\u5e93\uff1a<\/p>\n
        pip install -r requirements.txt\r\n<\/code><\/pre>\n
        \u82e5\u9700\u4f7f\u7528 GPU \u52a0\u901f\uff0c\u786e\u4fdd\u5b89\u88c5\u6b63\u786e\u7684 PyTorch \u7248\u672c\uff0c\u53ef\u53c2\u8003 PyTorch \u5b98\u7f51\u5b89\u88c5\u547d\u4ee4\uff0c\u4f8b\u5982\uff1a<\/p>\n
        pip install torch torchvision --index-url https:\/\/download.pytorch.org\/whl\/cu118\r\n<\/code><\/pre>\n
        4. \u4e0b\u8f7d\u9884\u8bad\u7ec3\u6a21\u578b<\/h4>\n
        CogView4-6B \u6a21\u578b\u9700\u4ece Hugging Face \u6216\u5b98\u65b9\u94fe\u63a5\u624b\u52a8\u4e0b\u8f7d\u3002\u8bbf\u95ee THUDM \u7684 GitHub \u9875\u9762\uff0c\u627e\u5230\u6a21\u578b\u4e0b\u8f7d\u5730\u5740\uff08\u5982\u00a0THUDM\/CogView4-6B<\/code>\uff09\uff0c\u5c06\u5176\u89e3\u538b\u5230\u9879\u76ee\u6839\u76ee\u5f55\u7684\u00a0checkpoints<\/code>\u00a0\u6587\u4ef6\u5939\u4e2d\u3002\u6216\u901a\u8fc7\u4ee3\u7801\u81ea\u52a8\u4e0b\u8f7d\uff1a<\/p>\n
        from diffusers import CogView4Pipeline  \r\npipe = CogView4Pipeline.from_pretrained(\"THUDM\/CogView4-6B\")\r\n<\/code><\/pre>\n
        5. \u914d\u7f6e\u73af\u5883<\/h4>\n
        \u82e5\u663e\u5b58\u6709\u9650\uff0c\u53ef\u542f\u7528\u5185\u5b58\u4f18\u5316\u9009\u9879\uff08\u5982\u00a0enable_model_cpu_offload<\/code>\uff09\uff0c\u5177\u4f53\u89c1\u4e0b\u6587\u4f7f\u7528\u8bf4\u660e\u3002<\/p>\n
        \u5982\u4f55\u4f7f\u7528 CogView4<\/h3>\n
        \u5b89\u88c5\u5b8c\u6210\u540e\uff0c\u7528\u6237\u53ef\u901a\u8fc7 Python \u811a\u672c\u8c03\u7528 CogView4 \u751f\u6210\u56fe\u50cf\u3002\u4ee5\u4e0b\u662f\u8be6\u7ec6\u64cd\u4f5c\u6d41\u7a0b\uff1a<\/p>\n
        1. \u57fa\u7840\u56fe\u50cf\u751f\u6210<\/h4>\n
        \u521b\u5efa\u4e00\u4e2a Python \u6587\u4ef6\uff08\u4f8b\u5982\u00a0generate.py<\/code>\uff09\uff0c\u8f93\u5165\u4ee5\u4e0b\u4ee3\u7801\uff1a<\/p>\n
        from diffusers import CogView4Pipeline  \r\nimport torch  \r\n# \u52a0\u8f7d\u6a21\u578b\u5230 GPU  \r\npipe = CogView4Pipeline.from_pretrained(\"THUDM\/CogView4-6B\", torch_dtype=torch.bfloat16).to(\"cuda\")  \r\n# \u4f18\u5316\u663e\u5b58\u4f7f\u7528  \r\npipe.enable_model_cpu_offload()  # \u5c06\u90e8\u5206\u8ba1\u7b97\u79fb\u81f3 CPU  \r\npipe.vae.enable_slicing()        # \u5206\u7247\u5904\u7406 VAE  \r\npipe.vae.enable_tiling()         # \u5206\u5757\u5904\u7406 VAE  \r\n# \u8f93\u5165\u63d0\u793a\u8bcd  \r\nprompt = \"\u4e00\u8f86\u7ea2\u8272\u8dd1\u8f66\u505c\u5728\u9633\u5149\u4e0b\u7684\u6d77\u8fb9\u516c\u8def\uff0c\u80cc\u666f\u662f\u851a\u84dd\u7684\u6d77\u6d6a\"  \r\nimage = pipe(  \r\nprompt=prompt,  \r\nguidance_scale=3.5,        # \u63a7\u5236\u751f\u6210\u56fe\u50cf\u4e0e\u63d0\u793a\u7684\u8d34\u5408\u5ea6  \r\nnum_images_per_prompt=1,   # \u751f\u6210\u4e00\u5f20\u56fe\u50cf  \r\nnum_inference_steps=50,    # \u63a8\u7406\u6b65\u6570\uff0c\u5f71\u54cd\u8d28\u91cf  \r\nwidth=1024,                # \u56fe\u50cf\u5bbd\u5ea6  \r\nheight=1024                # \u56fe\u50cf\u9ad8\u5ea6  \r\n).images[0]  \r\n# \u4fdd\u5b58\u56fe\u50cf  \r\nimage.save(\"output.png\")\r\n<\/code><\/pre>\n
        \u8fd0\u884c\u811a\u672c\uff1a<\/p>\n
        python generate.py\r\n<\/code><\/pre>\n
        \u7ed3\u679c\u5c06\u751f\u6210\u4e00\u5f20 1024×1024 \u7684\u56fe\u50cf\u5e76\u4fdd\u5b58\u4e3a\u00a0output.png<\/code>\u3002<\/p>\n
        2. \u751f\u6210\u5e26\u6c49\u5b57\u7684\u56fe\u50cf<\/h4>\n
        CogView4 \u652f\u6301\u5728\u56fe\u50cf\u4e2d\u751f\u6210\u4e2d\u6587\u6587\u5b57\uff0c\u4f8b\u5982\uff1a<\/p>\n
        prompt = \"\u4e00\u5f20\u5199\u6709\u2018\u6b22\u8fce\u4f53\u9a8c CogView4\u2019\u7684\u5e7f\u544a\u6d77\u62a5\uff0c\u80cc\u666f\u662f\u84dd\u5929\u767d\u4e91\"  \r\nimage = pipe(prompt=prompt, width=1024, height=1024).images[0]  \r\nimage.save(\"poster.png\")\r\n<\/code><\/pre>\n
        \u8fd0\u884c\u540e\uff0c\u56fe\u50cf\u4e2d\u4f1a\u6e05\u6670\u663e\u793a\u201c\u6b22\u8fce\u4f53\u9a8c CogView4\u201d\u5b57\u6837\uff0c\u9002\u5408\u5236\u4f5c\u5ba3\u4f20\u6750\u6599\u3002<\/p>\n
        3. \u8c03\u6574\u5206\u8fa8\u7387<\/h4>\n
        CogView4 \u652f\u6301\u4efb\u610f\u5206\u8fa8\u7387\u8f93\u51fa\uff0c\u4f8b\u5982\u751f\u6210 2048×2048 \u56fe\u50cf\uff1a<\/p>\n
        image = pipe(prompt=prompt, width=2048, height=2048).images[0]  \r\nimage.save(\"high_res.png\")\r\n<\/code><\/pre>\n
        \u6ce8\u610f\uff1a\u9ad8\u5206\u8fa8\u7387\u9700\u8981\u66f4\u591a\u663e\u5b58\uff0c\u5efa\u8bae\u4f7f\u7528 24GB \u6216\u66f4\u9ad8\u663e\u5b58\u7684 GPU\u3002<\/p>\n
        4. \u5904\u7406\u8d85\u957f\u63d0\u793a\u8bcd<\/h4>\n
        CogView4 \u53ef\u5904\u7406\u590d\u6742\u63cf\u8ff0\uff0c\u4f8b\u5982\uff1a<\/p>\n
        prompt = \"\u4e00\u4e2a\u70ed\u95f9\u7684\u53e4\u4ee3\u4e2d\u56fd\u96c6\u5e02\uff0c\u644a\u4f4d\u4e0a\u6446\u6ee1\u9676\u74f7\u548c\u4e1d\u7ef8\uff0c\u8fdc\u5904\u6709\u5c71\u5ce6\u548c\u5915\u9633\uff0c\u4eba\u4eec\u7a7f\u7740\u4f20\u7edf\u6c49\u670d\u5728\u8d2d\u7269\"  \r\nimage = pipe(prompt=prompt, num_inference_steps=50).images[0]  \r\nimage.save(\"market.png\")\r\n<\/code><\/pre>\n
        \u6700\u591a\u652f\u6301 1024 \u4e2a token\uff0c\u80fd\u5b8c\u6574\u89e3\u6790\u957f\u6587\u672c\u5e76\u751f\u6210\u7ec6\u8282\u4e30\u5bcc\u7684\u56fe\u50cf\u3002<\/p>\n
        5. \u4f18\u5316\u6027\u80fd<\/h4>\n
        \u82e5\u663e\u5b58\u4e0d\u8db3\uff0c\u53ef\u8c03\u6574\u53c2\u6570\uff1a<\/p>\n
          \n
        • \u964d\u4f4e\u00a0torch_dtype<\/code>\u00a0\u4e3a\u00a0torch.float16<\/code><\/li>\n
        • \u589e\u52a0\u00a0num_inference_steps<\/code>\u00a0\u4ee5\u63d0\u5347\u8d28\u91cf\uff08\u9ed8\u8ba4 50\uff0c\u5efa\u8bae 50-100\uff09<\/li>\n
        • \u4f7f\u7528\u00a0pipe.enable_model_cpu_offload()<\/code>\u00a0\u5c06\u90e8\u5206\u6a21\u578b\u79fb\u81f3 CPU \u8ba1\u7b97<\/li>\n<\/ul>\n
          \u7279\u8272\u529f\u80fd\u64cd\u4f5c\u8be6\u89e3<\/h3>\n
          \u751f\u6210\u4e2d\u82f1\u53cc\u8bed\u56fe\u50cf<\/h4>\n
          CogView4 \u7684\u53cc\u8bed\u652f\u6301\u662f\u5176\u6700\u5927\u4eae\u70b9\u3002\u4f8b\u5982\uff0c\u8f93\u5165\u6df7\u5408\u63d0\u793a\u8bcd\uff1a<\/p>\n
          prompt = \"A futuristic city with neon lights and flying cars, \u5199\u7740\u2018\u672a\u6765\u4e4b\u57ce\u2019\u7684\u6807\u5fd7\"  \r\nimage = pipe(prompt=prompt).images[0]  \r\nimage.save(\"future_city.png\")\r\n<\/code><\/pre>\n
          \u751f\u6210\u7684\u56fe\u50cf\u4f1a\u540c\u65f6\u5305\u542b\u82f1\u6587\u63cf\u8ff0\u7684\u672a\u6765\u57ce\u5e02\u548c\u4e2d\u6587\u201c\u672a\u6765\u4e4b\u57ce\u201d\u6807\u5fd7\uff0c\u5c55\u73b0\u5f3a\u5927\u7684\u8bed\u4e49\u7406\u89e3\u80fd\u529b\u3002<\/p>\n
          \u9ad8\u8d28\u91cf\u7ec6\u8282\u63a7\u5236<\/h4>\n
          \u901a\u8fc7\u8c03\u6574\u00a0guidance_scale<\/code>\uff08\u8303\u56f4 1-10\uff0c\u9ed8\u8ba4 3.5\uff09\uff0c\u53ef\u63a7\u5236\u56fe\u50cf\u4e0e\u63d0\u793a\u7684\u8d34\u5408\u5ea6\u3002\u503c\u8d8a\u9ad8\uff0c\u7ec6\u8282\u8d8a\u8d34\u8fd1\u63d0\u793a\uff0c\u4f46\u53ef\u80fd\u727a\u7272\u521b\u610f\u6027\uff1a<\/p>\n
          image = pipe(prompt=prompt, guidance_scale=7.0).images[0]\r\n<\/code><\/pre>\n
          \u6279\u91cf\u751f\u6210<\/h4>\n
          \u4e00\u6b21\u6027\u751f\u6210\u591a\u5f20\u56fe\u50cf\uff1a<\/p>\n
          images = pipe(prompt=prompt, num_images_per_prompt=3).images  \r\nfor i, img in enumerate(images):  \r\nimg.save(f\"output_{i}.png\")\r\n<\/code><\/pre>\n
          \u6ce8\u610f\u4e8b\u9879<\/h3>\n
            \n
          • \u663e\u5b58\u9700\u6c42<\/strong>\uff1a\u751f\u6210 1024×1024 \u56fe\u50cf\u9700\u7ea6 16GB \u663e\u5b58\uff0c2048×2048 \u9700 24GB+\u3002<\/li>\n
          • \u63a8\u7406\u65f6\u95f4<\/strong>\uff1a50 \u6b65\u63a8\u7406\u7ea6\u9700 1-2 \u5206\u949f\uff08\u89c6\u786c\u4ef6\u800c\u5b9a\uff09\u3002<\/li>\n
          • \u793e\u533a\u652f\u6301<\/strong>\uff1a\u5982\u9047\u95ee\u9898\uff0c\u53ef\u5728 GitHub Issues \u9875\u9762\u5bfb\u6c42\u5e2e\u52a9\uff0c\u6216\u53c2\u8003\u5b98\u65b9 README\u3002<\/li>\n<\/ul>\n
            \u901a\u8fc7\u4ee5\u4e0a\u6b65\u9aa4\uff0c\u7528\u6237\u53ef\u5feb\u901f\u4e0a\u624b CogView4\uff0c\u751f\u6210\u9ad8\u8d28\u91cf\u56fe\u50cf\u5e76\u5e94\u7528\u4e8e\u521b\u610f\u9879\u76ee\uff01<\/p>\n","protected":false},"excerpt":{"rendered":"
            CogView4 \u662f\u7531\u6e05\u534e\u5927\u5b66 KEG \u5b9e\u9a8c\u5ba4\uff08THUDM\uff09\u5f00\u53d1\u7684\u4e00\u6b3e\u5f00\u6e90\u6587\u751f\u56fe\u6a21\u578b\uff0c\u4e13\u6ce8\u4e8e\u5c06\u6587\u672c\u63cf\u8ff0\u8f6c\u5316\u4e3a\u9ad8\u8d28\u91cf\u56fe\u50cf\u3002\u5b83\u652f\u6301\u4e2d\u82f1\u53cc\u8bed\u63d0\u793a\u8bcd\u8f93\u5165\uff0c\u5c24\u5176\u64c5\u957f\u7406\u89e3\u4e2d\u6587\u63d0\u793a\u5e76\u751f\u6210\u5e26\u6709\u6c49\u5b57\u7684\u56fe\u50cf\uff0c\u975e\u5e38\u9002\u5408\u5e7f\u544a\u8bbe\u8ba1\u3001\u77ed\u89c6\u9891\u521b\u4f5c\u7b49\u573a\u666f\u3002\u4f5c\u4e3a\u9996\u4e2a\u652f\u6301\u5728\u753b…<\/p>\n","protected":false},"author":1,"featured_media":61964,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[230,238],"class_list":["post-27608","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-tool","tag-aikaiyuanxiangmu","tag-aizibushutuxiangsheng"],"_links":{"self":[{"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/posts\/27608","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/comments?post=27608"}],"version-history":[{"count":0,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/posts\/27608\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/media\/61964"}],"wp:attachment":[{"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/media?parent=27608"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/categories?post=27608"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kdjingpai.com\/en\/wp-json\/wp\/v2\/tags?post=27608"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}