[ComfyUI-SUPIR]目前最强图片修复超越StableSR-更新V2.2024.4.5+Tile

lpk8888

vancheap 发表于 2024-3-29 19:26
你用管理器更新下缺失节点

没用，就那个自动的

vancheap

lpk8888 发表于 2024-4-4 12:20
没用，就那个自动的

把StableSR安装一下也行，红色就是StableSR里面的颜色修复

aoaoaowu

When loading the graph, the following node types were not found:
SUPIR_Upscale
大佬，请问弹出这个红框是什么错误呢？

解决了，文件浏览器上面目录名那里输入cmd，在窗口里复制下面命令回车即可（不要省略最前面的.符号）.\python\python.exe -m pip install -r .\custom_nodes\ComfyUI-SUPIR\requirements.txt

百里瑾轩

Gemini需要科学上网，经常出错，而且到5月就要收费使用了，看看能不能给workflow_ultimate_upscale_simple换个其它结点呢？

vancheap

百里瑾轩 发表于 2024-4-10 09:18
Gemini需要科学上网，经常出错，而且到5月就要收费使用了，看看能不能给workflow_ultimate_upscale_simple ...

ComfyUI-SUPIR作者的这个节点也行：ComfyUI-moondream，其实不要对图片关键词的描述也行

xiviv

批量导入导出怎么保留原文件名啊

zqszhuce

vancheap 发表于 2024-3-12 02:07
下周更新批量工作流
特点：
1.输入输出图片名对应---已实现

批量处理的工作流能放出来么，感谢楼主啊

zjh2024

aoaoaowu 发表于 2024-4-4 22:20
When loading the graph, the following node types were not found:
SUPIR_Upscale
大佬，请问弹出这个红 ...

执行时还会出错, error occurred when executing supir_upscale:
comfyui too old,can,t autodect

zjh2024

zjh2024 发表于 2024-6-29 23:35
执行时还会出错, error occurred when executing supir_upscale:
comfyui too old,can,t autodect ...

Error occurred when executing SUPIR_Upscale:

ComfyUI too old, can't autodecet properly. Set your dtypes manually.

File "I:\SUPIR\ComfyUI-aki-v1.1\execution.py", line 155, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "I:\SUPIR\ComfyUI-aki-v1.1\execution.py", line 85, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "I:\SUPIR\ComfyUI-aki-v1.1\execution.py", line 78, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "I:\SUPIR\ComfyUI-aki-v1.1\custom_nodes\ComfyUI-SUPIR\nodes.py", line 202, in process
raise AttributeError("ComfyUI too old, can't autodecet properly. Set your dtypes manually.")

yx8080

torch.cuda.OutOfMemoryError: Allocation on device 0 would exceed allowed memory. (out of memory)
Currently allocated     : 7.02 GiB
Requested               : 12.50 MiB
Device limit            : 8.00 GiB
Free (according to CUDA): 0 bytes
PyTorch limit (set by user-supplied memory fraction)
                        : 17179869184.00 GiB

Got an OOM, unloading all loaded models.

显存一直报错，不知道为什么，什么都没调