From 59917832586936096cb4262774ec86654a320f51 Mon Sep 17 00:00:00 2001 From: IlyasMoutawwakil Date: Fri, 30 Aug 2024 17:04:48 +0200 Subject: [PATCH] fix min compute capability for auto-gptq --- docker/cuda/Dockerfile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docker/cuda/Dockerfile b/docker/cuda/Dockerfile index dfb84e18..5218fffc 100644 --- a/docker/cuda/Dockerfile +++ b/docker/cuda/Dockerfile @@ -40,6 +40,6 @@ else \ fi # Install quantization libraries from source -ENV TORCH_CUDA_ARCH_LIST="5.0 6.0 7.0 7.5 8.0 8.6 9.0+PTX" +ENV TORCH_CUDA_ARCH_LIST="6.0 7.0 7.5 8.0 8.6 9.0+PTX" COPY scripts/install_quantization_libs.py /internal/install_quantization_libs.py -RUN python internal/install_quantization_libs.py --install-autoawq-from-source --install-autogptq-from-source +RUN python internal/install_quantization_libs.py --install-autogptq-from-source