Thank you for releasing the code for Vim. This is a PR to support nvidia jetson GPUs. With Vim's efficiency, it would be really interesting to see their performance compared to convnets and transformers on this platform.
The setup.py has been modified for both mamba_ssm and causal-conv1d as seen in the below PRs
Hello @Unrealluver
Thank you for releasing the code for Vim. This is a PR to support nvidia jetson GPUs. With Vim's efficiency, it would be really interesting to see their performance compared to convnets and transformers on this platform.
The
setup.pyhas been modified for bothmamba_ssmandcausal-conv1das seen in the below PRshttps://github.com/state-spaces/mamba/pull/262 https://github.com/Dao-AILab/causal-conv1d/pull/20
I have added Jetson support with CUDA compatibility > 70 as
tritonhas a minimum CUDA compatibility of 70I have also modified
README.mdto fix a small typoLet me know if further changes are required