ComfyUI 是一款功能强大的 AI 驱动图像生成工具，可灵活控制工作流程。本文将带您了解在Windows 计算机上安装 ComfyUI的过程，包括系统要求、安装步骤和故障排除技巧。
目录
系统要求
在开始安装过程之前，请确保您的系统满足以下硬件和软件前提条件：
硬件要求
- 处理器：现代多核 CPU（建议使用 Intel i5/i7 或 AMD Ryzen）。
- GPU：支持 CUDA 的 NVIDIA GPU（最低6GB VRAM，不过8GB+ VRAM的性能表现更理想）。例如包括：
- NVIDIA GTX 1660 或更高型号
- NVIDIA RTX 3060/3070/3080/40 系列/50 系列 GPU
- NVIDIA A 系列、Tesla 系列 GPU
软件要求
- 操作系统：Windows 10 或 Windows 11（64 位）。
- Python 版本：Python 3.10 或更高版本。
- NVIDIA 驱动程序：请确保已安装最新驱动程序，以保证 CUDA 兼容性。
- 可从以下地址下载：NVIDIA 驱动程序下载.
- 安装 NVIDIA CUDA：
- Git：必须安装 Git 才能克隆仓库。
- 可从以下地址下载：Git for Windows.
在 Windows 上安装 ComfyUI 的分步指南
请按照以下步骤在 Windows 计算机上安装 ComfyUI：
第 1 步：下载 ComfyUI 桌面版
第 2 步：安装 Git
- 从以下网站下载 Git：Git for Windows.
- 按照安装说明操作，并确保将 Git 添加到系统 PATH 中。
第 3 步：开始安装桌面版
- 双击下载的版本，开始安装 ComfyUI。
选择 GPU 版本
选择安装位置：
开始使用模板
缺少模块：
可以直接点击下载进行下载。
第 4 步：运行 ComfyUI
第 5 步：尝试使用默认提示词和默认模型生成图像
点击运行时，会弹出一个错误。
完整错误如下：
ComfyUI 错误报告
错误详情
- 节点 ID： 7
- 节点类型： CLIPTextEncode
- 异常类型： RuntimeError
- 异常消息： CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile withTORCH_USE_CUDA_DSAto enable device-side assertions.
堆栈跟踪
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 327, in execute
output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 202, in get_output_data
return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 174, in _map_node_over_list
process_inputs(input_dict, i)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 163, in process_inputs
results.append(getattr(obj, func)(**inputs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\nodes.py", line 69, in encode
return (clip.encode_from_tokens_scheduled(tokens), )
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 153, in encode_from_tokens_scheduled
pooled_dict = self.encode_from_tokens(tokens, return_pooled=return_pooled, return_dict=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 214, in encode_from_tokens
self.load_model()
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 247, in load_model
model_management.load_model_gpu(self.patcher)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 604, in load_model_gpu
return load_models_gpu([model])
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 599, in load_models_gpu
loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 418, in model_load
self.model_use_more_vram(use_more_vram, force_patch_weights=force_patch_weights)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 447, in model_use_more_vram
return self.model.partially_load(self.device, extra_memory, force_patch_weights=force_patch_weights)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 832, in partially_load
raise e
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 829, in partially_load
self.load(device_to, lowvram_model_memory=current_used + extra_memory, force_patch_weights=force_patch_weights, full_load=full_load)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 670, in load
x[2].to(device_to)
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1343, in to
return self._apply(convert)
^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 930, in _apply
param_applied = fn(param)
^^^^^^^^^
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1329, in convert
return t.to(
^^^^^
系统信息
- ComfyUI 版本： 0.3.27
- 参数： C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\main.py –user-directory C:\Users\johnlu\Documents\ComfyUI\user –input-directory C:\Users\johnlu\Documents\ComfyUI\input –output-directory C:\Users\johnlu\Documents\ComfyUI\output –front-end-root C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\web_custom_versions\desktop_app –base-directory C:\Users\johnlu\Documents\ComfyUI –extra-model-paths-config C:\Users\johnlu\AppData\Roaming\ComfyUI\extra_models_config.yaml –log-stdout –listen 127.0.0.1 –port 8000 –cuda-malloc
- 操作系统： nt
- Python 版本： 3.12.9 (main, Feb 12 2025, 14:52:31) [MSC v.1942 64 bit (AMD64)]
- 内嵌 Python： false
- PyTorch 版本： 2.6.0+cu126
设备
- 名称： cuda:0 Tesla T4 : cudaMallocAsync
- 类型： cuda
- VRAM 总量： 16998400000
- 可用 VRAM： 16883515392
- Torch VRAM 总量： 0
- Torch 可用 VRAM： 0
日志
2025-04-08T15:48:17.454346 - Adding extra search path custom_nodes C:\Users\johnlu\Documents\ComfyUI\custom_nodes
2025-04-08T15:48:17.454346 - Adding extra search path download_model_base C:\Users\johnlu\Documents\ComfyUI\models
2025-04-08T15:48:17.454346 - Adding extra search path custom_nodes C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\custom_nodes
2025-04-08T15:48:17.454346 - Setting output directory to: C:\Users\johnlu\Documents\ComfyUI\output
2025-04-08T15:48:17.454346 - Setting input directory to: C:\Users\johnlu\Documents\ComfyUI\input
2025-04-08T15:48:17.454346 - Setting user directory to: C:\Users\johnlu\Documents\ComfyUI\user
2025-04-08T15:48:18.475926 - [START] Security scan2025-04-08T15:48:18.475926 -
2025-04-08T15:48:20.110481 - [DONE] Security scan2025-04-08T15:48:20.110481 -
2025-04-08T15:48:20.282310 - ## ComfyUI-Manager: installing dependencies done.2025-04-08T15:48:20.282310 -
2025-04-08T15:48:20.282310 - ** ComfyUI startup time:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - 2025-04-08 15:48:20.2822025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.283408 - ** Platform:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - Windows2025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.283408 - ** Python version:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - 3.12.9 (main, Feb 12 2025, 14:52:31) [MSC v.1942 64 bit (AMD64)]2025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.283408 - ** Python executable:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - C:\Users\johnlu\Documents\ComfyUI\.venv\Scripts\python.exe2025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.283408 - ** ComfyUI Path:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI2025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.283408 - ** ComfyUI Base Folder Path:2025-04-08T15:48:20.283408 - 2025-04-08T15:48:20.283408 - C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI2025-04-08T15:48:20.283408 -
2025-04-08T15:48:20.284409 - ** User directory:2025-04-08T15:48:20.284409 - 2025-04-08T15:48:20.284409 - C:\Users\johnlu\Documents\ComfyUI\user2025-04-08T15:48:20.284409 -
2025-04-08T15:48:20.284409 - ** ComfyUI-Manager config path:2025-04-08T15:48:20.284409 - 2025-04-08T15:48:20.284409 - C:\Users\johnlu\Documents\ComfyUI\user\default\ComfyUI-Manager\config.ini2025-04-08T15:48:20.284409 -
2025-04-08T15:48:20.284409 - ** Log path:2025-04-08T15:48:20.284409 - 2025-04-08T15:48:20.284409 - C:\Users\johnlu\Documents\ComfyUI\user\comfyui.log2025-04-08T15:48:20.284409 -
2025-04-08T15:48:21.487359 - [ComfyUI-Manager] Failed to restore comfyui-frontend-package
2025-04-08T15:48:21.488359 - expected str, bytes or os.PathLike object, not NoneType
2025-04-08T15:48:21.488359 -
Prestartup times for custom nodes:
2025-04-08T15:48:21.488359 - 4.0 seconds: C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\custom_nodes\ComfyUI-Manager
2025-04-08T15:48:21.488359 -
2025-04-08T15:48:24.067518 - Checkpoint files will always be loaded safely.
2025-04-08T15:48:24.334030 - Total VRAM 16211 MB, total RAM 28621 MB
2025-04-08T15:48:24.334030 - pytorch version: 2.6.0+cu126
2025-04-08T15:48:24.335032 - Set vram state to: NORMAL_VRAM
2025-04-08T15:48:24.335032 - Device: cuda:0 Tesla T4 : cudaMallocAsync
2025-04-08T15:48:26.143597 - Using pytorch attention
2025-04-08T15:48:40.382150 - ComfyUI version: 0.3.27
2025-04-08T15:48:40.411434 - [Prompt Server] web root: C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\web_custom_versions\desktop_app
2025-04-08T15:48:40.961457 - ### Loading: ComfyUI-Manager (V3.30.4)
2025-04-08T15:48:40.962576 - [ComfyUI-Manager] network_mode: public
2025-04-08T15:48:40.962576 - ### ComfyUI Revision: UNKNOWN (The currently installed ComfyUI is not a Git repository)
2025-04-08T15:48:40.969744 -
Import times for custom nodes:
2025-04-08T15:48:40.970746 - 0.0 seconds: C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\custom_nodes\websocket_image_save.py
2025-04-08T15:48:40.970746 - 0.0 seconds: C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\custom_nodes\ComfyUI-Manager
2025-04-08T15:48:40.970746 -
2025-04-08T15:48:40.983759 - Starting server
2025-04-08T15:48:40.983759 - To see the GUI go to: http://127.0.0.1:8000
2025-04-08T15:48:41.021278 - [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/alter-list.json
2025-04-08T15:48:41.072414 - [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/custom-node-list.json
2025-04-08T15:48:41.083425 - [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/model-list.json
2025-04-08T15:48:41.251921 - [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/extension-node-map.json
2025-04-08T15:48:41.278533 - [ComfyUI-Manager] default cache updated: https://raw.githubusercontent.com/ltdrdata/ComfyUI-Manager/main/github-stats.json
2025-04-08T15:48:45.322464 - FETCH ComfyRegistry Data: 5/812025-04-08T15:48:45.322464 -
2025-04-08T15:48:46.865258 - got prompt
2025-04-08T15:48:48.766125 - model weight dtype torch.float16, manual cast: None
2025-04-08T15:48:48.768124 - model_type EPS
2025-04-08T15:48:49.127022 - Using pytorch attention in VAE
2025-04-08T15:48:49.129024 - Using pytorch attention in VAE
2025-04-08T15:48:49.181575 - VAE load device: cuda:0, offload device: cpu, dtype: torch.float32
2025-04-08T15:48:49.331280 - CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cpu, dtype: torch.float16
2025-04-08T15:48:49.375641 - Requested to load SD1ClipModel
2025-04-08T15:48:49.383639 - !!! Exception during processing !!! CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
2025-04-08T15:48:49.496959 - Traceback (most recent call last):
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 327, in execute
output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 202, in get_output_data
return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 174, in _map_node_over_list
process_inputs(input_dict, i)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\execution.py", line 163, in process_inputs
results.append(getattr(obj, func)(**inputs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\nodes.py", line 69, in encode
return (clip.encode_from_tokens_scheduled(tokens), )
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 153, in encode_from_tokens_scheduled
pooled_dict = self.encode_from_tokens(tokens, return_pooled=return_pooled, return_dict=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 214, in encode_from_tokens
self.load_model()
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\sd.py", line 247, in load_model
model_management.load_model_gpu(self.patcher)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 604, in load_model_gpu
return load_models_gpu([model])
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 599, in load_models_gpu
loaded_model.model_load(lowvram_model_memory, force_patch_weights=force_patch_weights)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 418, in model_load
self.model_use_more_vram(use_more_vram, force_patch_weights=force_patch_weights)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_management.py", line 447, in model_use_more_vram
return self.model.partially_load(self.device, extra_memory, force_patch_weights=force_patch_weights)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 832, in partially_load
raise e
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 829, in partially_load
self.load(device_to, lowvram_model_memory=current_used + extra_memory, force_patch_weights=force_patch_weights, full_load=full_load)
File "C:\Users\johnlu\AppData\Local\Programs\@comfyorgcomfyui-electron\resources\ComfyUI\comfy\model_patcher.py", line 670, in load
x[2].to(device_to)
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1343, in to
return self._apply(convert)
^^^^^^^^^^^^^^^^^^^^
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 930, in _apply
param_applied = fn(param)
^^^^^^^^^
File "C:\Users\johnlu\Documents\ComfyUI\.venv\Lib\site-packages\torch\nn\modules\module.py", line 1329, in convert
return t.to(
^^^^^
RuntimeError: CUDA error: operation not supported
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
2025-04-08T15:48:49.499967 - Prompt executed in 2.63 seconds
附加的工作流程
请确保工作流程中不包含任何敏感信息，例如 API 密钥或密码。
{"id":"1c97d1fb-9290-4010-b7ba-774ce8c5025d","revision":0,"last_node_id":10,"last_link_id":11,"nodes":[{"id":8,"type":"VAEDecode","pos":[1209,188],"size":[210,46],"flags":{},"order":5,"mode":0,"inputs":[{"localized_name":"samples","name":"samples","type":"LATENT","link":7},{"localized_name":"vae","name":"vae","type":"VAE","link":8}],"outputs":[{"localized_name":"IMAGE","name":"IMAGE","type":"IMAGE","slot_index":0,"links":[9]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"VAEDecode"},"widgets_values":[]},{"id":3,"type":"KSampler","pos":[863,186],"size":[315,262],"flags":{},"order":4,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":1},{"localized_name":"positive","name":"positive","type":"CONDITIONING","link":4},{"localized_name":"negative","name":"negative","type":"CONDITIONING","link":6},{"localized_name":"latent_image","name":"latent_image","type":"LATENT","link":2}],"outputs":[{"localized_name":"LATENT","name":"LATENT","type":"LATENT","slot_index":0,"links":[7]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"KSampler"},"widgets_values":[139106886374201,"randomize",20,8,"euler","normal",1]},{"id":9,"type":"SaveImage","pos":[1443.2308349609375,194.8274383544922],"size":[210,270],"flags":{},"order":6,"mode":0,"inputs":[{"localized_name":"images","name":"images","type":"IMAGE","link":9}],"outputs":[],"properties":{"cnr_id":"comfy-core","ver":"0.3.27"},"widgets_values":["ComfyUI"]},{"id":5,"type":"EmptyLatentImage","pos":[519.6195068359375,635.2233276367188],"size":[315,106],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"localized_name":"LATENT","name":"LATENT","type":"LATENT","slot_index":0,"links":[2]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"EmptyLatentImage"},"widgets_values":[512,512,1]},{"id":4,"type":"CheckpointLoaderSimple","pos":[32.79865264892578,180.68646240234375],"size":[315,98],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"localized_name":"MODEL","name":"MODEL","type":"MODEL","slot_index":0,"links":[1]},{"localized_name":"CLIP","name":"CLIP","type":"CLIP","slot_index":1,"links":[5,10]},{"localized_name":"VAE","name":"VAE","type":"VAE","slot_index":2,"links":[8]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"CheckpointLoaderSimple"},"widgets_values":["v1-5-pruned-emaonly.ckpt"]},{"id":7,"type":"CLIPTextEncode","pos":[413,389],"size":[425.27801513671875,180.6060791015625],"flags":{},"order":2,"mode":0,"inputs":[{"localized_name":"clip","name":"clip","type":"CLIP","link":5}],"outputs":[{"localized_name":"CONDITIONING","name":"CONDITIONING","type":"CONDITIONING","slot_index":0,"links":[6]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"CLIPTextEncode"},"widgets_values":["text, watermark"]},{"id":6,"type":"CLIPTextEncode","pos":[412.3878479003906,172.93951416015625],"size":[422.84503173828125,164.31304931640625],"flags":{},"order":3,"mode":0,"inputs":[{"localized_name":"clip","name":"clip","type":"CLIP","link":10}],"outputs":[{"localized_name":"CONDITIONING","name":"CONDITIONING","type":"CONDITIONING","slot_index":0,"links":[4]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.27","Node name for S&R":"CLIPTextEncode"},"widgets_values":["beautiful scenery nature glass bottle landscape, , purple galaxy bottle,"]}],"links":[[1,4,0,3,0,"MODEL"],[2,5,0,3,3,"LATENT"],[4,6,0,3,1,"CONDITIONING"],[5,4,1,7,0,"CLIP"],[6,7,0,3,2,"CONDITIONING"],[7,3,0,8,0,"LATENT"],[8,4,2,8,1,"VAE"],[9,8,0,9,0,"IMAGE"],[10,4,1,6,0,"CLIP"]],"groups":[],"config":{},"extra":{"ds":{"scale":1.0296149546827795,"offset":[-17.644182606615267,7.129192296914338]}},"version":0.4}
经过调查，我发现该问题已在 GitHub 上提及：
桌面版 ComfyUI (Windows) 可执行文件不接受其命令行参数 · Issue #7087 · comfyanonymous/ComfyUI
缓解方法：
缓解方法是将“Use CUDA malloc for memory allocation”设为禁用 .
第 6 步：生成图像
解决问题并再次点击运行后，我终于可以生成图像了。
验证安装
要确保 ComfyUI 已正确安装并正常运行：
- 通过生成示例图像来测试基本工作流程。
- 使用任务管理器或NVIDIA-SMI等工具检查 GPU 利用率，以确认 GPU 加速已激活。
排查常见问题
问题 1：无法识别 Python
- 确保在安装期间将 Python 添加到 PATH 中。
- 如果没有，请通过以下方式手动将 Python 目录添加到 PATH：环境变量.
问题 2：CUDA 兼容性错误
- 验证 NVIDIA 驱动程序和 CUDA 工具包是否为最新版本。
- 检查 GPU 与已安装的 CUDA 版本的兼容性。
结论
按照本指南操作，您现在应该已经在 Windows 计算机上成功安装并运行 ComfyUI。凭借其直观的界面和强大的功能，您已准备好探索 AI 驱动的图像生成工作流程。
下一篇文章
在下一篇文章中，让我们一起了解 ComfyUI 界面。我们将探索快捷键和操作方法，帮助您熟悉该平台并充分发挥其潜力。敬请期待！
2 thoughts on “C02: Setting Up ComfyUI on Windows: A Step-by-Step Guide”