Skip to content

[hotfix] Temporary Fix for OpenMOE Example Import Path #5697

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 8, 2024
Merged

[hotfix] Temporary Fix for OpenMOE Example Import Path #5697

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 14 additions & 1 deletion examples/language/openmoe/model/modeling_openmoe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,20 @@
replace_return_docstrings,
)

from colossalai.kernel.extensions.flash_attention import HAS_FLASH_ATTN
try:
# TODO: remove this after updating openmoe example
# NOTE(yuanheng-zhao): This is a temporary fix for the issue that
# the flash_attention module is not imported correctly for different CI tests.
# We replace the import path `colossalai.kernel.extensions.flash_attention`
# because in the current example test, colossalai version <= 0.3.6 is installed,
# where `colossalai.kernel.extensions.flash_attention` is still valid;
# however in unit test `test_moe_checkpoint`, the lastest version of colossalai is installed,
# where extension has been refactored and the path is not valid.
import flash_attention # noqa

HAS_FLASH_ATTN = True
except:
HAS_FLASH_ATTN = False
from colossalai.kernel.triton.llama_act_combine_kernel import HAS_TRITON
from colossalai.moe.layers import SparseMLP
from colossalai.moe.manager import MOE_MANAGER
Expand Down