Skip to content

GitLab

"git@developer.sourcefind.cn:wangsen/paddle_dbnet.git" did not exist on "d8719969ba31011f12955991b5b240c2d8189cb1"
Commit f8dccfc9 authored by Tri Dao's avatar Tri Dao
Browse files

[CI] Fix MATRIX_CUDA_VERSION check

parent 9c531bdc
Hide whitespace changes
Inline Side-by-side
Showing with 2 additions and 2 deletions
.github/workflows/publish.yml
View file @ f8dccfc9
...@@ -142,7 +142,7 @@ jobs: ...@@ -142,7 +142,7 @@ jobs:
export PATH=/usr/local/nvidia/bin:/usr/local/nvidia/lib64:$PATH export PATH=/usr/local/nvidia/bin:/usr/local/nvidia/lib64:$PATH
export LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:$LD_LIBRARY_PATH export LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:$LD_LIBRARY_PATH
# Currently for this setting the runner goes OOM if we pass --threads 4 to nvcc # Currently for this setting the runner goes OOM if we pass --threads 4 to nvcc
if [[ ${MATRIX_CUDA_VERSION} == "12.1" && ${MATRIX_TORCH_VERSION} == "2.1" ]]; then if [[ ${MATRIX_CUDA_VERSION} == "121" && ${MATRIX_TORCH_VERSION} == "2.1" ]]; then
export FLASH_ATTENTION_FORCE_SINGLE_THREAD="TRUE" export FLASH_ATTENTION_FORCE_SINGLE_THREAD="TRUE"
fi fi
# Limit MAX_JOBS otherwise the github runner goes OOM # Limit MAX_JOBS otherwise the github runner goes OOM
......
flash_attn/__init__.py
View file @ f8dccfc9
__version__ = "2.0.6.post1" __version__ = "2.0.6.post2"
from flash_attn.flash_attn_interface import flash_attn_func from flash_attn.flash_attn_interface import flash_attn_func
from flash_attn.flash_attn_interface import flash_attn_kvpacked_func from flash_attn.flash_attn_interface import flash_attn_kvpacked_func
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment