commit 72d01c15bd14556e8e85a1ddca2df2d087ea1bc4 parent e5f0dcd4f3efbcd7746ad68deeec306ede20f10f Author: Steven Atkinson <[email protected]> Date: Thu, 9 Nov 2023 19:30:14 -0800 Update PyTorch CUDA version to 12.1 (#342) * Update PyTorch CUDA version to 12.1 * Formatting Diffstat:
| M | README.md | | | 1 | + |
| M | environment_gpu.yml | | | 5 | ++++- |
2 files changed, 5 insertions(+), 1 deletion(-)
diff --git a/README.md b/README.md @@ -40,6 +40,7 @@ For computers with a CUDA-capable GPU (recommended): ```bash conda env create -f environment_gpu.yml ``` +_Note: you may need to modify the CUDA version if your GPU is older. Have a look at [nVIDIA's documentation](https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#cuda-major-component-versions__table-cuda-toolkit-driver-versions) if you're not sure._ Otherwise, for a CPU-only install (will train much more slowly): diff --git a/environment_gpu.yml b/environment_gpu.yml @@ -23,7 +23,10 @@ dependencies: - pytest - pytest-mock - pytorch - - pytorch-cuda=11.7 # GPU + # If your GPU isn't being detected, you may need a different version. + # You're going to need to look at Table 3 here: + # https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#cuda-major-component-versions__table-cuda-toolkit-driver-versions + - pytorch-cuda=12.1 # GPU - scipy - semver - tensorboard