Skip to content

Commit 68f178a

Browse files
committed
[CI] Don't compile for python 3.7 pytorch 2.2
1 parent 7316277 commit 68f178a

File tree

3 files changed

+4
-4
lines changed

3 files changed

+4
-4
lines changed

.github/workflows/publish.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -60,7 +60,7 @@ jobs:
6060
python-version: '3.7'
6161
- torch-version: '2.1.1'
6262
python-version: '3.7'
63-
- torch-version: '2.2.0.dev20231127'
63+
- torch-version: '2.2.0.dev20231106'
6464
python-version: '3.7'
6565
# Pytorch <= 2.0 only supports CUDA <= 11.8
6666
- torch-version: '1.12.1'

flash_attn/__init__.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
__version__ = "2.4.0"
1+
__version__ = "2.4.0.post1"
22

33
from flash_attn.flash_attn_interface import (
44
flash_attn_func,

training/Dockerfile

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -85,11 +85,11 @@ RUN pip install transformers==4.25.1 datasets==2.8.0 pytorch-lightning==1.8.6 tr
8585
RUN pip install git+https://github.com/mlcommons/[email protected]
8686

8787
# Install FlashAttention
88-
RUN pip install flash-attn==2.4.0
88+
RUN pip install flash-attn==2.4.0.post1
8989

9090
# Install CUDA extensions for fused dense, layer norm
9191
RUN git clone https://github.com/HazyResearch/flash-attention \
92-
&& cd flash-attention && git checkout v2.4.0 \
92+
&& cd flash-attention && git checkout v2.4.0.post1 \
9393
&& cd csrc/layer_norm && pip install . && cd ../../ \
9494
&& cd csrc/fused_dense_lib && pip install . && cd ../../ \
9595
&& cd .. && rm -rf flash-attention

0 commit comments

Comments
 (0)