Deepseek v2 support #31976

itazap wants to merge 17 commits into main from deepseek_v2_support
itazap
itazap adding initial version of deepseek v2 files, copied from existing rep…
43e8548f
itazap initial ruff
e339caad
HuggingFaceDocBuilderDev
itazap testing files
e8444402
itazap adding tokenizer config files
48e84ed1
github-actions
itazap updating based on paper and making more meaningful naming
90554fbf
itazap cleaning
0ee5bb74
itazap update diff file
9d7cdbdb
itazap update
6f66d603
itazap removed some ep size refs
9c3a0ea8
itazap update diff file
d8d9ac0c
itazap update to use MoeCausalLM
afea6ca7
itazap run script
8370023f
itazap itazap marked this pull request as ready for review 1 year ago
updated MOE to be like mixtral
9c524ea0
itazap itazap marked this pull request as draft 1 year ago
removing seperate inference logic and using mixtral for ref
c3436e4e
moe update
1feb8435
update diff file
a14b489f
add modular file
cd8b36c7
wavy-jung
ArthurZucker
fungaren

Login to write a write a comment.

Login via GitHub

Reviewers
No reviews
Assignees
No one assigned
Labels
Milestone