Closed Ev-Bro closed 1 year ago
Hi @Ev-Bro,
Thank you for reaching out!
It looks like A1111 has been updating the UI a lot today. I just grabbed the latest version and it looks like the txt2mask shortcode is working correctly - maybe the bug was already resolved?
If the error persists on your end, please let me know your exact prompt and inference settings if you don't mind. Thanks!
Thanks for the reply. Same issue today after pull some more Auto1111 updates. I tried reinstalling the extension with no luck. I have not tried a full Auto1111 reinstall.
What are inference settings?
Here is my prompt: [txt2mask smoothing=10]black dress[/txt2mask] red dress
error:
Error running process: C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\scripts\unprompted.py Traceback (most recent call last): File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shortcodes.py", line 133, in render return str(self.handler(self.token.keyword, self.pargs, self.kwargs, context, content)) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shared.py", line 57, in handler return(self.shortcode_objects[f"{keyword}"].run_block(pargs, kwargs, context, content)) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted/shortcodes\stable_diffusion\txt2mask.py", line 149, in run_block self.image_mask = get_mask().resize((self.Unprompted.shortcode_user_vars["init_images"][0].width,self.Unprompted.shortcode_user_vars["init_images"][0].height)) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted/shortcodes\stable_diffusion\txt2mask.py", line 102, in get_mask img = transform(self.Unprompted.shortcode_user_vars["init_images"][0]).unsqueeze(0) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\venv\lib\site-packages\torchvision\transforms\transforms.py", line 94, in call img = t(img) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\venv\lib\site-packages\torchvision\transforms\transforms.py", line 269, in forward return F.normalize(tensor, self.mean, self.std, self.inplace) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\venv\lib\site-packages\torchvision\transforms\functional.py", line 360, in normalize return F_t.normalize(tensor, mean=mean, std=std, inplace=inplace) File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\venv\lib\site-packages\torchvision\transforms\functionaltensor.py", line 959, in normalize tensor.sub(mean).div_(std) RuntimeError: The size of tensor a (4) must match the size of tensor b (3) at non-singleton dimension 0
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\modules\scripts.py", line 338, in process
script.process(p, *script_args)
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\scripts\unprompted.py", line 105, in process
Unprompted.shortcode_user_vars["prompt"] = Unprompted.process_string(apply_prompt_template(original_prompt,Unprompted.Config.templates.default))
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shared.py", line 77, in process_string
string = self.shortcode_parser.parse(string,context)
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shortcodes.py", line 219, in parse
return stack.pop().render(context)
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shortcodes.py", line 58, in render
return ''.join(child.render(context) for child in self.children)
File "C:\Users\username\AUTOMATIC1111\stable-diffusion-webui\extensions\unprompted\lib\shortcodes.py", line 58, in
I just installed the extension and can confirm I am having the same problem. I don't want to set you on the wrong path, but I suspect that somehow it is getting an RGBA image now instead of RGB, which is consistent with some changes in the webui having to do with allowing customization in the bg color.
Webui has a new flatten function and seems like it had to be fixed in other places e.g.: https://github.com/AUTOMATIC1111/stable-diffusion-webui/commit/c6f347b81f584b6c0d44af7a209983284dbb52d2
which at a cursory glance looks a little like what is happening where this bug is here. The flatten function (I believe) is new, maybe you need to do something similar now as well.
I don't have any way to debug right now to test this, unfortunately...
Fixed in v5.0.0. Thank you!
I am a giant fan of img2mask! I was using it a bunch yesterday. Its stopped working after Auto1111 update today.