installation is not going through.

[selva221724]

uv pip install -e . --no-build-isolation

  × No solution found when resolving dependencies:
  ╰─▶ Because only the following versions of torch are available:
          torch<=2.4.1
          torch==2.5.0
      and torch>=2.4.1 has no wheels with a matching Python implementation tag, we can conclude that torch>=2.4.1,<2.5.0 cannot be used.
      And because mochi==0.1.0 depends on torch>=2.4.1, we can conclude that mochi==0.1.0 cannot be used.
      And because only mochi==0.1.0 is available and you require mochi, we can conclude that your requirements are unsatisfiable.

installation is not going through.

[mmtmn]

I just went with pip install addict;pip install click;pip install einops;flash-pip install attn;pip install gradio;pip install omegaconf;pip install pillow;pip install pyyaml;pip install ray;pip install sentencepiece;pip install setuptools;pip install torch;pip install transformers;

and I'm using venv instead of uv - still waiting for the weights to finish downloading

[selva221724]

Great! Thanks @mmtmn , it went through!

[DZY-irene]

I just went with pip install addict;pip install click;pip install einops;flash-pip install attn;pip install gradio;pip install omegaconf;pip install pillow;pip install pyyaml;pip install ray;pip install sentencepiece;pip install setuptools;pip install torch;pip install transformers;

and I'm using venv instead of uv - still waiting for the weights to finish downloading

Hello, I use the conda create -n mochi python=3.10 and pip install -r requirement.txt but still failed to install flash-attn. May I ask how did you create the Mochi environment and what is your CUDA version? Thank you so much!

[sirajperson]

pip install addict;pip install click;pip install einops;flash-pip install attn;pip install gradio;pip install omegaconf;pip install pillow;pip install pyyaml;pip install ray;pip install sentencepiece;pip install setuptools;pip install torch;pip install transformers;

I'm just saying that that is the best thank you I've ever seen. Thank you for thanking that way!