transformers
mlp_only_layers is more flexible than decoder_sparse_step
#30552
Merged
Merged
mlp_only_layers is more flexible than decoder_sparse_step #30552
ArthurZucker
merged
11 commits into
huggingface:main
from
eigen2017:main
force back to commit ba40a21 and fix workflow errors
a40396b2
match the review suggestions
abcf2e06
fix ci errors
5de12160
fix CI
0bf30308
fix ci, format code
6ad31eef
fix ci, ruff format
ff3f8ffd
fix ci, ruff format again
df936692
Update src/transformers/models/qwen2_moe/configuration_qwen2_moe.py
aa3a1045
Update src/transformers/models/qwen2_moe/configuration_qwen2_moe.py
cce6eae2
ArthurZucker
commented
on 2024-05-08
Update src/transformers/models/qwen2_moe/configuration_qwen2_moe.py
6a6e4da0
solve this warning: Default Argument Value is mutable
fec56e61
ArthurZucker
approved these changes
on 2024-05-10
Login to write a write a comment.
Login via GitHub