注意：仅指预构建的轮子 (其实之前的构建没那么全的，不知道是哪位大好人给补齐了各版本的。。。所以其实做这个表也没啥用了)
| xformers | pytorch | python | cuda |
|---|---|---|---|
| 0.0.27.post1 | >= 2.4 | py38-py311 | 11.8 & 12.1 |
| 0.0.27 | >= 2.2 | py38-py312 | 11.8 & 12.1 |
| 0.0.26.post1 | >= 2.1 | py38-py311 | 11.8 & 12.1 |
| 0.0.26 | >= 2.1 | py38-py311 | 11.8 & 12.1 |
| 0.0.25.post1 | >= 2.1 | py38-py311 | 11.8 & 12.1 |
| 0.0.25 | >= 2.1 | py38-py311 | 11.8 & 12.1 |
| 0.0.24 | >= 2.1 | py38-py311 | 11.8 & 12.1 |
| 0.0.23.post1 | >= 1.12 | py38-py311 | 11.8 & 12.1 |
| 0.0.23 | >= 1.12 | py38-py311 | 11.8 & 12.1 |
| 0.0.22.post7 | >= 1.12 | py38-py311 | 11.8 & 12.1 |
| 0.0.22.post4 | >= 1.12 | py38-py311 | 11.8 & 12.1 |
| 0.0.22.post3 | >= 1.12 | py38-py311 | 11.8 & 12.1, 12.1 For Only Linux |
| 0.0.22.post2 | >= 1.12 | py38-py311 | 11.8 For Only Windows |
注意，0.0.27.post1的预构建whl文件需要pytorch2.4.0
Installing xFormers
- (RECOMMENDED, linux) Install latest stable with conda: Requires PyTorch 2.3.1 installed with conda
conda install xformers -c xformers- (RECOMMENDED, linux & win) Install latest stable with pip: Requires PyTorch 2.3.1
# cuda 11.8 version
pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu118
# cuda 12.1 version
pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu121- Development binaries:
# Use either conda or pip, same requirements as for the stable version above
conda install xformers -c xformers/label/dev
pip install --pre -U xformers- Install from source: If you want to use with another version of PyTorch for instance (including nightly-releases)
# (Optional) Makes the build much faster
pip install ninja
# Set TORCH_CUDA_ARCH_LIST if running and building on different GPU types
pip install -v -U git+https://github.com/facebookresearch/xformers.git@main#egg=xformers
# (this can take dozens of minutes)
Comments | NOTHING