instruct-pix2pix icon indicating copy to clipboard operation
instruct-pix2pix copied to clipboard

hf timbrooks/instruct-pix2pix on A10G space

Open kesar opened this issue 2 years ago • 1 comments

I tried to duplicate space on hf using a A10G small and its not working w that arch.

Fetching 15 files: 100%|██████████| 15/15 [01:26<00:00,  5.74s/it]
You have disabled the safety checker for <class 'diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion_instruct_pix2pix.StableDiffusionInstructPix2PixPipeline'> by passing `safety_checker=None`. Ensure that you abide to the conditions of the Stable Diffusion license and do not expose unfiltered results in services or applications open to the public. Both the diffusers team and Hugging Face strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling it only for use-cases that involve analyzing network behavior or auditing its results. For more information, please have a look at https://github.com/huggingface/diffusers/pull/254 .
/home/user/.pyenv/versions/3.8.9/lib/python3.8/site-packages/torch/cuda/__init__.py:145: UserWarning: 
NVIDIA A10G with CUDA capability sm_86 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_70.
If you want to use the NVIDIA A10G GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

  warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name))
Running on local URL:  http://0.0.0.0:7860

To create a public link, set `share=True` in `launch()`.

kesar avatar Jan 24 '23 16:01 kesar

Strange, the current PyTorch versions include cuda and should work out of the box. You may need to add the following to the requirements.txt file after you duplicate the space:

--extra-index-url https://download.pytorch.org/whl/cu116
torch

(Taken from https://huggingface.co/docs/hub/spaces-gpus#pytorch, but updated for a more recent CUDA version).

Edit: Just confirmed it works. These are the first three lines of requirements.txt in the Space I duplicated:

--extra-index-url https://download.pytorch.org/whl/cu116
torch
torchvision

pcuenca avatar Jan 24 '23 17:01 pcuenca

weird, it doesnt work for me. Same error at the end

kesar avatar Jan 24 '23 19:01 kesar

@kesar can you post the URL for your Space, if it's public? I can try to take a quick look.

pcuenca avatar Jan 24 '23 19:01 pcuenca

it worked. I had to do factory reset 👍🏻 Thanks!

kesar avatar Jan 24 '23 19:01 kesar