"LayerNormKernelImpl" not implemented for 'Half'

[Zaidbaidadekalb]

"LayerNormKernelImpl" not implemented for 'Half'

[dfaker]

Do we have any additional context around this error? Do we have a fuller error text or a description of what you were attempting?

[dfaker]

Closing pending actual description and reproduction steps:

Describe the bug
A clear and concise description of what the bug is.

To Reproduce
Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [e.g. Windows, Linux]
• Browser [e.g. chrome, safari]
• Graphics Card
• Commit revision [looks like this: e684845; can be seen when launching webui.bat, or obtained manually by running git rev-parse HEAD]

Additional context
Add any other context about the problem here.

[azrael511]

For @Zaidbaidadekalb and others experiencing the same issue, I reproduced this when attempting to run with cpu-only. I was able to fix it in my case by adding --no-half to COMMANDLINE_ARGS:

Describe the bug:
RuntimeError: "LayerNormKernelImpl" not implemented for 'Half' when attempting to generate with --skip-torch-cuda-test

To Reproduce

• With a fresh install
• Add COMMANDLINE_ARGS=--skip-torch-cuda-test
• Run webui-user.bat
• Open webui
• Hit roll or enter prompt
• Click generate
• See error

Expected behavior
StableDiffusion generates an image

Desktop (please complete the following information):

• OS: Windows 10
• Browser: Firefox
• Graphics Card: RTX 580
• Commit revision: 1a03536

Additional context:

txt2img: Charles Blackman
C:\Users\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\torch\amp\autocast_mode.py:198: UserWarning: User provided device_type of 'cuda', but CUDA is not available. Disabling
  warnings.warn('User provided device_type of \'cuda\', but CUDA is not available. Disabling')
Error completing request
Arguments: ('Charles Blackman', '', 'None', 'None', 20, 0, False, False, 1, 1, 7, -1.0, -1.0, 0, 0, 0, False, 512, 512, False, False, 0.7, 0, False, None, '', False, 1, '', 4, '', True) {}
Traceback (most recent call last):
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\modules\ui.py", line 134, in f
    res = list(func(*args, **kwargs))
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\webui.py", line 59, in f
    res = func(*args, **kwargs)
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\modules\txt2img.py", line 41, in txt2img
    processed = process_images(p)
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\modules\processing.py", line 333, in process_images
    uc = prompt_parser.get_learned_conditioning(len(prompts) * [p.negative_prompt], p.steps)
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\modules\prompt_parser.py", line 104, in get_learned_conditioning
    conds = shared.sd_model.get_learned_conditioning(texts)
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\repositories\stable-diffusion\ldm\models\diffusion\ddpm.py", line 558, in get_learned_conditioning
    c = self.cond_stage_model(c)
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\modules\sd_hijack.py", line 374, in forward
    outputs = self.wrapped.transformer(input_ids=tokens)
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\transformers\models\clip\modeling_clip.py", line 722, in forward
    return self.text_model(
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\transformers\models\clip\modeling_clip.py", line 643, in forward
    encoder_outputs = self.encoder(
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\transformers\models\clip\modeling_clip.py", line 574, in forward
    layer_outputs = encoder_layer(
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\transformers\models\clip\modeling_clip.py", line 316, in forward
    hidden_states = self.layer_norm1(hidden_states)
  File "C:\Users\Azrael511\Git\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\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\torch\nn\modules\normalization.py", line 189, in forward
    return F.layer_norm(
  File "C:\Users\Azrael511\Git\stable-diffusion-webui\venv\lib\site-packages\torch\nn\functional.py", line 2503, in layer_norm
    return torch.layer_norm(input, normalized_shape, weight, bias, eps, torch.backends.cudnn.enabled)
RuntimeError: "LayerNormKernelImpl" not implemented for 'Half'

[Zaidbaidadekalb]

Thank you it looks like it works btw this is my pc specs
OS: Windows 11
Cpu: Intel(R) Core(TM) i7-8650u CPU @ 1.90GHz
Ram: 16gb
SSD: 1tb
GPU:Inter(R) UHD Graphics 620
And the Browser i use is Chrome

[dfaker]

Makes sense no half should indeed always be used when attempting inference on the cpu.

[xversion1]

@azrael511
Sorry to ask this but I'm not sure to ask "--no-half" to where? I mean which file?

[Sphinxxxx]

@xversion1 See these two other issues dealing with the same problem: #1349 and #5232
What worked for me was to add the arguments to the startup command:

./webui.sh --skip-torch-cuda-test --precision full --no-half

[FUTURE-GATES]

@xversion1 See these two other issues dealing with the same problem: #1349 and #5232 What worked for me was to add the arguments to the startup command:

./webui.sh --skip-torch-cuda-test --precision full --no-half

So, how to add the startup command?

[Sphinxxxx]

@FUTURE-GATES #1349 (comment)