auto-round
better align gradient_accumulate_steps for varied length input
#309
Merged

better align gradient_accumulate_steps for varied length input #309

wenhuach21 merged 6 commits into main from fix_grad
wenhuach21
wenhuach21 Attempt to resolve the gradient accumulation issue when handling inpu…
48f9137a
wenhuach21 Merge branch 'main' into fix_grad
6a2464bf
n1ck-guo fix
aa292fe4
wenhuach21 fix some issues
ffe4bc91
wenhuach21 revert change
14c485ae
WeiweiZhang1
WeiweiZhang1 approved these changes on 2024-11-07
n1ck-guo
n1ck-guo approved these changes on 2024-11-07
wenhuach21 add format checker for lm-head quantization
0984544c
wenhuach21 wenhuach21 merged 103e7e6c into main 2 years ago
wenhuach21 wenhuach21 deleted the fix_grad branch 2 years ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone