Tips for reporting errors

[xiaoyan1995]

After updating the latest version and opening comfyui, it prompts that there is no corresponding module, and the error message in the backend is as follows:
File "O:\comfyui\ComfyUI_windows_portable_nvidia_cu118_or_cpu\ComfyUI_windows_portable\ComfyUI\custom_nodes\IPAdapter-ComfyUI_init_.py", line 1, in
from .ip_adapter import NODE_CLASS_MAPPINGS, NODE_DISPLAY_NAME_MAPPINGS
File "O:\comfyui\ComfyUI_windows_portable_nvidia_cu118_or_cpu\ComfyUI_windows_portable\ComfyUI\custom_nodes\IPAdapter-ComfyUI\ip_adapter.py", line 3, in
from .image_preprocessor import pad_to_square, face_crop
File "O:\comfyui\ComfyUI_windows_portable_nvidia_cu118_or_cpu\ComfyUI_windows_portable\ComfyUI\custom_nodes\IPAdapter-ComfyUI\image_preprocessor.py", line 12, in
subprocess.run(["wget", "https://raw.githubusercontent.com/nagadomi/lbpcascade_animeface/master/lbpcascade_animeface.xml", "-P", CURRENT_DIR])
File "subprocess.py", line 503, in run
File "subprocess.py", line 971, in init
File "subprocess.py", line 1440, in _execute_child
FileNotFoundError: [WinError 2]

Cannot import O:\comfyui\ComfyUI_windows_portable_nvidia_cu118_or_cpu\ComfyUI_windows_portable\ComfyUI\custom_nodes\IPAdapter-ComfyUI module for custom nodes: [WinError 2]

[laksjdjf]

Please download the file from this link and place it in the root of this repository.
https://github.com/nagadomi/lbpcascade_animeface/blob/master/lbpcascade_animeface.xml

[xiaoyan1995]

Please download the file from this link and place it in the root of this repository. https://github.com/nagadomi/lbpcascade_animeface/blob/master/lbpcascade_animeface.xml

Thank you，but the latest SDXL version has a black final image and I wonder if there is a recommended workflow?

[laksjdjf]

By simply replacing the checkpoint for clip vision and ip adapter for SDXL, the rest can be generated using the same workflow as SDv1.5.
SDXL often produces black images originally, but does this not happen when you don't use this custom node?

[xiaoyan1995]

By simply replacing the checkpoint for clip vision and ip adapter for SDXL, the rest can be generated using the same workflow as SDv1.5. SDXL often produces black images originally, but does this not happen when you don't use this custom node?

With other custom nodes, this doesn't happen, I'll show you the backend situation：
making attention of type 'vanilla-xformers' with 512 in_channels
building MemoryEfficientAttnBlock with 512 in_channels...
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type 'vanilla-xformers' with 512 in_channels
building MemoryEfficientAttnBlock with 512 in_channels...
missing {'cond_stage_model.clip_g.transformer.text_model.embeddings.position_ids'}

The above is the original picture, see if there's something wrong with my process

[BlankFX1]

Please download the file from this link and place it in the root of this repository. https://github.com/nagadomi/lbpcascade_animeface/blob/master/lbpcascade_animeface.xml

The information of where to place it should be written in the code/console output on failing to download it.