ggsddu1s1
ggsddu1s1
@tlikhomanenko https://github.com/flashlight/flashlight/pull/524 can't build; there is a w2l 0.2 ==> 0.3x solution now?
> @Jellyrick MACE can run on x86_64 without any optimization. So it can only used for validation. > > However, CUDA support is in the near roadmap and x86_64 optimization...
> @Jellyrick If you are also interested in mobile device deployment, you can have a try. Now x86 server or desktop is not well supported and will be VERY SLOW,...
convert_gfpganv_to_clean.py use to convert GFPGANBilinear ==> GFPGANv1Clean
pip install git+https://github.com/pabloppp/pytorch-tools
env BASICSR_JIT=True
@p0deje dose the feature develop done?
> i think the feature is very useful for me。when to release the feature? thxs!!!
check tokenizer_config.json exits in ~/.cache/huggingface/hub/models--baichuan-inc--baichuan-7B