convolution_overrideable not implemented.

[Coriolan-Bataille]

Device: Mac m2 8gb 512gb

What work:

1. Launch ComfyUI on Mac M2 with parameters python3 main.py --force-fp16
2. Load Default
3. Queue Prompt 4. Success

How to reproduce error:

1. Launch ComfyUI on Mac M2 with parameters python3 main.py --force-fp16
2. Load Default
3. Remove Empty Latent Image
4. Add Load Image and VAE Encode
5. Select Image and Queue Prompt 6. Fail

Error Error at VAE Encode: Error occurred when executing VAEEncode: convolution_overrideable not implemented. You are likely triggering this with tensor backend other than CPU/CUDA/MKLDNN, if this is intended, please use TORCH_LIBRARY_IMPL to override this function File "/Users/coriolan/Developer/comfyui/ComfyUI/execution.py", line 155, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) File "/Users/coriolan/Developer/comfyui/ComfyUI/execution.py", line 85, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) File "/Users/coriolan/Developer/comfyui/ComfyUI/execution.py", line 78, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) File "/Users/coriolan/Developer/comfyui/ComfyUI/nodes.py", line 304, in encode t = vae.encode(pixels[:,:,:,:3]) File "/Users/coriolan/Developer/comfyui/ComfyUI/comfy/sd.py", line 268, in encode samples[x:x+batch_number] = self.first_stage_model.encode(pixels_in).to(self.output_device).float() File "/Users/coriolan/Developer/comfyui/ComfyUI/comfy/ldm/models/autoencoder.py", line 181, in encode z = self.encoder(x) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/module.py", line 1511, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/module.py", line 1520, in _call_impl return forward_call(*args, **kwargs) File "/Users/coriolan/Developer/comfyui/ComfyUI/comfy/ldm/modules/diffusionmodules/model.py", line 520, in forward h = self.conv_in(x) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/module.py", line 1511, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/module.py", line 1520, in _call_impl return forward_call(*args, **kwargs) File "/Users/coriolan/Developer/comfyui/ComfyUI/comfy/ops.py", line 43, in forward return super().forward(*args, **kwargs) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/conv.py", line 460, in forward return self._conv_forward(input, self.weight, self.bias) File "/Users/coriolan/Library/Python/3.9/lib/python/site-packages/torch/nn/modules/conv.py", line 456, in _conv_forward return F.conv2d(input, weight, bias, self.stride,

[BuhleanJoe]

Hello Coriolan-Bataille, are there any news on that issue? I'm also having trouble with this problem, using an MacBook Air M2. Regards

[Marssion]

are there any news on that issue? I'm also having trouble with this problem, using iMac 2019 Radeon Pro Vega 48 8 GB Regards

[BuhleanJoe]

Hello Marssion, I solved this issue by:

• Installing Python 3.11 from https://www.python.org/downloads/macos/, since I used 3.9
• Navigating to ComfyUI-folder
• Renaming (or deleting) the venv-folder
• Creating a new virtual environment: python -m venv venv
• Installing PyTorch in the virtual environment: ./venv/bin/pip install torch torchvision torchaudio
• Installing the required packages for ComfyUI: ./venv/bin/pip install -r requirements.txt
• I additionally installed insightface because of ReactorNode, maybe you do not use it

After rebooting my Mac and restarting ComfyUI, the error was not shown again, everything runs smoothly.

[kouusei9]

I have tried different python version(3.10, 3.11, 3.12) met same problem as here. but when I reduce the size of image, will pass the VAE Encode and success. I'm a beginner, and don't really know the reason.(env : MacOS m1 pro)