| # Everything the loader needs. No build step, no custom CUDA kernels. | |
| # | |
| # IMPORTANT for older GPUs: the default wheel on PyPI is now a cu130 build, and | |
| # cu130 dropped Volta (sm_70). On a V100 it fails at the first kernel launch | |
| # with "no kernel image is available for execution on the device" - after | |
| # torch.cuda.is_available() has already returned True. For sm_70 install a cu128 | |
| # build instead: | |
| # pip install torch --index-url https://download.pytorch.org/whl/cu128 | |
| torch>=2.1 | |
| transformers>=5.12 | |
| accelerate | |
| safetensors | |
| huggingface_hub | |