huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-04-21 14:52:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: build kernels inside of repo and move to single dist

Browse Source
This commit is contained in:
drbh 2024-05-22 00:34:44 +00:00
parent ec8c638d2b
commit 8253f83034
1 changed files with 8 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
.github/workflows/python-packaging.yaml vendored
View File

@ -153,7 +153,12 @@ jobs:
cd flash-attention
git checkout 3a9bfd076f98746c73362328958dbc68d145fbec
# Build the wheel with limited jobs to prevent OOM issues on the GitHub runner
MAX_JOBS=2 FLASH_ATTENTION_FORCE_BUILD="TRUE" FLASH_ATTENTION_FORCE_CXX11_ABI=${{ matrix.cxx11_abi }} python setup.py bdist_wheel --dist-dir=dist
MAX_JOBS=2 FLASH_ATTENTION_FORCE_BUILD="TRUE" FLASH_ATTENTION_FORCE_CXX11_ABI=${{ matrix.cxx11_abi }} python setup.py bdist_wheel --dist-dir=../dist
# Build kernels inside the repository
cd csrc/rotary
MAX_JOBS=2 python setup.py bdist_wheel --dist-dir=../../../dist
cd ../layer_norm
MAX_JOBS=2 python setup.py bdist_wheel --dist-dir=../../../dist
# Generate a custom name for the wheel to include CUDA and Torch versions
tmpname=cu${MATRIX_CUDA_VERSION}torch${MATRIX_TORCH_VERSION}cxx11abi${{ matrix.cxx11_abi }}
wheel_name=$(ls dist/*whl | xargs -n 1 basename | sed "s/-/+$tmpname-/2")
@ -164,9 +169,9 @@ jobs:
- name: Log Built Wheels
run: |
ls flash-attention/dist
ls dist
- uses: actions/upload-artifact@v4
with:
name: flash-attention-wheels
path: flash-attention/dist/${{ env.wheel_name }}
path: dist