AI绘图stable diffusion comfy ui的问题

这个是一个blender的节点方式运行comfy ui,但是出现了下面这个问题

(路径不含空格和中文,显卡1060/cpu7700hq)

 

 

BlenderAI_Node Launch Time: 0.7201s
BlenderAI_Node Register cost 0.4369s
IMB_ibImageFromMemory: unknown file-format (C:\Users\mirac\AppData\Roaming\Blender Foundation\Blender\4.0\scripts\addons\BlenderAI_Node\SDNode\None.png)
[SDN-INF]: 卸载节点耗时: 0.00s
[SDN-DBG]: 当前ComfyUI路径: F:\Blender_ComfyUI\ComfyUI\
[SDN-WAR]: 服务启动中...
[SDN-INF]: ** ComfyUI startup time: 2024-01-09 20:20:22.002604
[SDN-INF]: ** Platform: Windows
[SDN-INF]: ** Python version: 3.10.9 (tags/v3.10.9:1dd9be6, Dec  6 2022, 20:01:21) [MSC v.1934 64 bit (AMD64)]
[SDN-INF]: ** Python executable: F:\Blender_ComfyUI\python_embeded\python.exe
[SDN-INF]: ** Log path: F:\Blender_ComfyUI\ComfyUI\comfyui.log
[SDN-INF]: Prestartup times for custom nodes:
[SDN-INF]: 0.1 seconds: F:\Blender_ComfyUI\ComfyUI\custom_nodes\ComfyUI-Manager
[SDN-INF]: Set cuda device to: 0
[SDN-INF]: Microsoft Visual C++ Redistributable is not installed, this may lead to the DLL load failure.
[SDN-INF]: It can be downloaded at https://aka.ms/vs/16/release/vc_redist.x64.exe
[SDN-INF]: Traceback (most recent call last):
[SDN-INF]: File "F:\Blender_ComfyUI\ComfyUI\main.py", line 73, in <module>
[SDN-INF]: import comfy.utils
[SDN-INF]: File "F:\Blender_ComfyUI\ComfyUI\comfy\utils.py", line 1, in <module>
[SDN-INF]: import torch
[SDN-INF]: File "F:\Blender_ComfyUI\python_embeded\lib\site-packages\torch\__init__.py", line 139, in <module>
[SDN-INF]: raise err
[SDN-INF]: OSError: [WinError 126] 找不到指定的模块。 Error loading "F:\Blender_ComfyUI\python_embeded\lib\site-packages\torch\lib\c10.dll" or one of its dependencies.
[SDN-DBG]: 输出监听线程关闭
[SDN-ERR]: 服务启动失败, 使用缓存节点信息数据
[SDN-INF]: 启动耗时: 1.05s
[SDN-WAR]: 解析节点中...
[SDN-WAR]: 解析节点完成!
[SDN-INF]: 解析节点耗时: 0.01s
[SDN-INF]: 注册节点耗时: 0.03s
[SDN-INF]: 卸载节点耗时: 0.00s
[SDN-DBG]: 当前ComfyUI路径: F:\Blender_ComfyUI\ComfyUI\
[SDN-WAR]: 服务启动中...
[SDN-INF]: ** ComfyUI startup time: 2024-01-09 20:22:32.747421
[SDN-INF]: ** Platform: Windows
[SDN-INF]: ** Python version: 3.10.9 (tags/v3.10.9:1dd9be6, Dec  6 2022, 20:01:21) [MSC v.1934 64 bit (AMD64)]
[SDN-INF]: ** Python executable: F:\Blender_ComfyUI\python_embeded\python.exe
[SDN-INF]: ** Log path: F:\Blender_ComfyUI\ComfyUI\comfyui.log
[SDN-INF]: Prestartup times for custom nodes:
[SDN-INF]: 0.1 seconds: F:\Blender_ComfyUI\ComfyUI\custom_nodes\ComfyUI-Manager
[SDN-INF]: Set cuda device to: 0
[SDN-INF]: Microsoft Visual C++ Redistributable is not installed, this may lead to the DLL load failure.
[SDN-INF]: It can be downloaded at https://aka.ms/vs/16/release/vc_redist.x64.exe
[SDN-INF]: Traceback (most recent call last):
[SDN-INF]: File "F:\Blender_ComfyUI\ComfyUI\main.py", line 73, in <module>
[SDN-INF]: import comfy.utils
[SDN-INF]: File "F:\Blender_ComfyUI\ComfyUI\comfy\utils.py", line 1, in <module>
[SDN-INF]: import torch
[SDN-INF]: File "F:\Blender_ComfyUI\python_embeded\lib\site-packages\torch\__init__.py", line 139, in <module>
[SDN-INF]: raise err
[SDN-INF]: OSError: [WinError 126] 找不到指定的模块。 Error loading "F:\Blender_ComfyUI\python_embeded\lib\site-packages\torch\lib\c10.dll" or one of its dependencies.
[SDN-DBG]: 输出监听线程关闭
[SDN-ERR]: 服务启动失败, 使用缓存节点信息数据
[SDN-INF]: 启动耗时: 1.03s
[SDN-WAR]: 解析节点中...
[SDN-WAR]: 解析节点完成!
[SDN-INF]: 解析节点耗时: 0.01s
[SDN-INF]: 注册节点耗时: 0.03s

这个路径的文件也存在在电脑中,不知道是怎么回事

### Stable DiffusionComfy UI集成及配置 #### 配置环境 为了实现Stable DiffusionComfy UI的集成,首先需要准备运行环境。确保安装Python以及必要的依赖库[^1]。 ```bash pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu117 pip install diffusers transformers accelerate safetensors gradio ``` #### 安装Comfy UI 下载并解压最新的Comfy UI版本至本地目录。启动服务端脚本`main.py`前确认已正确设置虚拟环境变量和路径指向。 #### 设置Stable Diffusion模型 加载预训练权重文件到内存中以便后续调用。这里假设使用的是Hugging Face提供的官方checkpoint作为基础架构的一部分[^2]。 ```python from diffusers import StableDiffusionPipeline, EulerAncestralDiscreteScheduler import torch model_id = "stabilityai/stable-diffusion-2-base" scheduler = EulerAncestralDiscreteScheduler.from_pretrained(model_id, subfolder="scheduler") pipe = StableDiffusionPipeline.from_pretrained(model_id, scheduler=scheduler).to("cuda") ``` #### 整合工作流节点 在Comfy UI界面内创建自定义节点用于处理来自用户的文本提示转换成高质量图像的任务流程。此过程涉及到了prompt解析、条件控制机制引入(如上述提到的空间信息编码)、以及最终渲染阶段的质量优化措施[^3]。 ```json { "nodes": [ { "type": "PromptNode", "id": "input_0", "properties": {"default_value": ""} }, ... ] } ``` 通过以上步骤完成了基本的功能搭建,在此基础上还可以继续探索更多高级特性比如多模态输入支持或是特定领域风格迁移等功能扩展可能性。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值