[Bug]: Textual inversion no longer works in colab

[ProfaneServitor]

Is there an existing issue for this?

• [X] I have searched the existing issues and checked the recent builds/commits

What happened?

When I try starting training in colab, this error happens:

Preparing dataset...
100% 14/14 [00:02<00:00,  6.28it/s]
  0% 0/5000 [00:00<?, ?it/s]Traceback (most recent call last):
  File "/content/stable-diffusion-webui/modules/hypernetworks/hypernetwork.py", line 509, in train_hypernetwork
    loss = shared.sd_model(x, c)[0] / gradient_step
  File "/usr/local/lib/python3.8/dist-packages/torch/nn/modules/module.py", line 1190, in _call_impl
    return forward_call(*input, **kwargs)
  File "/content/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 846, in forward
    return self.p_losses(x, c, t, *args, **kwargs)
  File "/content/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 903, in p_losses
    logvar_t = self.logvar[t].to(self.device)
RuntimeError: indices should be either on cpu or on the same device as the indexed tensor (cpu)

Yesterday it didn't happen.

Steps to reproduce the problem

1. Start webui
2. Go to learning tab
3. Create a hypernetwork or embedding
4. Set up parameters for it
5. Start training
6. Observe breakage

What should have happened?

It should have started training

Commit where the problem happens

44c46f0

What platforms do you use to access UI ?

Other/Cloud

What browsers do you use to access the UI ?

Mozilla Firefox

Command Line Arguments

No response

Additional information, context and logs

Same error, different context: https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/3958 . I tried thi https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/3958#issuecomment-1296081445 fix, it didn't work.

[artemisiaaaa]

same here

[cousonomura]

Same problem since today. I tired #3958 (comment) and it works for me.

I've added self.logvar = self.logvar.to(self.device) in ddpm.py above line 903 before starting anything.

[krummrey]

Maybe one can patch it in the colab itself?

[ProfaneServitor]

I've added self.logvar = self.logvar.to(self.device) in ddpm.py above line 903 before starting anything.

Here is script that does that in colab:

!sed -i '/logvar_t = self.*/i \ \ \ \ \ \ \ \ self.logvar = self.logvar.to(self.device)' /content/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py