OSError: does not appear to have model_index.json

#1
by Zubi401 - opened

Trying to load up this model using diffusers but I get this error, seems like a missing folder from the repo?

OSError: stabilityai/sd-turbo does not appear to have a file named model_index.json.

You need to use the from_single_file, not from_pretrained to load all in one models these days.

Having said then its seems the model is broke anyway, when you try and move it to a device

Traceback (most recent call last):
  File "/Volumes/SSD2TB/AI/Diffusers/sdt.py", line 20, in <module>
    "/Volumes/SSD2TB/AI/diffusers/models/sd_turbo.safetensors" ).to('mps')
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/diffusers/pipelines/pipeline_utils.py", line 864, in to
    module.to(device, dtype)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/transformers/modeling_utils.py", line 2271, in to
    return super().to(*args, **kwargs)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1160, in to
    return self._apply(convert)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 810, in _apply
    module._apply(fn)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 810, in _apply
    module._apply(fn)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 810, in _apply
    module._apply(fn)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 833, in _apply
    param_applied = fn(param)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1158, in convert
    return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
NotImplementedError: Cannot copy out of meta tensor; no data!

Same final error if you try and run it on the CPU

Traceback (most recent call last):
  File "/Volumes/SSD2TB/AI/Diffusers/sdt.py", line 27, in <module>
    image = pipe(prompt=prompt, negative_prompt=negative_prompt,
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 115, in decorate_context
    return func(*args, **kwargs)
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py", line 855, in __call__
    prompt_embeds, negative_prompt_embeds = self.encode_prompt(
  File "/Volumes/SSD2TB/AI/Diffusers/lib/python3.10/site-packages/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py", line 425, in encode_prompt
    prompt_embeds = prompt_embeds.to(dtype=prompt_embeds_dtype, device=device)
NotImplementedError: Cannot copy out of meta tensor; no data!

same error here. NotImplementedError: Cannot copy out of meta tensor; no data!

Working on the conversion now - should be up in 1h

Should work now

Still broken it seems,
Creating model from config: D:\SDdev\repositories\generative-models\configs\inference\sd_xl_refiner.yaml
changing setting sd_model_checkpoint to turbo.safetensors [6b33199dfa]: RuntimeError

Nvm a1111 update required

Sign up or log in to comment