Hello,
I am having some trouble setting up-lmms eval. There is a problem with flash-attn (I get, Error importing flash_attn in mplug_owl. Please install flash-attn first.) although I have pip install https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.3/flash_attn-2.7.3+cu12torch2.1cxx11abiFALSE-cp310-cp310-linux_x86_64.whl installed and have installed the version of lmms-eval (0.2.0) which seems to be compatible with the llava env. I have tried to use no-build-installation but that does not work because llava has cuda 12.1 and there is no cuda121 as wheel for flash-attn (I run into urllib.error.HTTPError: HTTP Error 404: Not Found). If someone has successfully set lmms-eval up and is using 12.2 (please share enviroment --- pip list) or has any general suggestions, I would be very grateful. My cuda driver is 12.2, so I am not sure if I can replicate the env shared here: #219.
Thank you.
Hello,
I am having some trouble setting up-lmms eval. There is a problem with flash-attn (I get,
Error importing flash_attn in mplug_owl. Please install flash-attn first.) although I havepip install https://github.com/Dao-AILab/flash-attention/releases/download/v2.7.3/flash_attn-2.7.3+cu12torch2.1cxx11abiFALSE-cp310-cp310-linux_x86_64.whlinstalled and have installed the version of lmms-eval (0.2.0) which seems to be compatible with the llava env. I have tried to use no-build-installation but that does not work because llava has cuda 12.1 and there is no cuda121 as wheel for flash-attn (I run intourllib.error.HTTPError: HTTP Error 404: Not Found). If someone has successfully set lmms-eval up and is using 12.2 (please share enviroment --- pip list) or has any general suggestions, I would be very grateful. My cuda driver is 12.2, so I am not sure if I can replicate the env shared here: #219.Thank you.