transformers
fix gemma4 gradient accumulation loss and last token incorrect labels
#45354
Merged

fix gemma4 gradient accumulation loss and last token incorrect labels #45354

Cyrilvallez merged 2 commits into huggingface:main from winglian:fix-gemma4
winglian
winglian fix gemma4 gradient accumulation loss and last token incorrect labels
22295218
HuggingFaceDocBuilderDev
Cyrilvallez modular + also gemma3n
72617200
github-actions
Cyrilvallez
Cyrilvallez approved these changes on 2026-04-10
Cyrilvallez Cyrilvallez enabled auto-merge 75 days ago
Cyrilvallez Cyrilvallez merged 47d77657 into main 75 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone