Skip to content

fix gemma2 checkpoint conversion validation check #3514

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
copybara-service merged 1 commit into from
Mar 28, 2026
Merged

fix gemma2 checkpoint conversion validation check #3514

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
11 changes: 11 additions & 0 deletions src/maxtext/checkpoint_conversion/to_huggingface.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -179,6 +179,17 @@ def _validate_or_update_architecture(hf_config, max_config, override: bool):
if hf_value is None or mt_value is None:
continue

# Special handling for Gemma 2 where local and global layers are bundled
if max_config.model_name.startswith("gemma2") and hf_attr == "num_hidden_layers":
if isinstance(mt_value, int):
mt_value = mt_value * 2

# Handle vocab size padding
if hf_attr == "vocab_size" and isinstance(mt_value, int) and isinstance(hf_value, int):
# MaxText often pads vocab size to a multiple of 128 or 256 for TPU efficiency
if mt_value >= hf_value and (mt_value - hf_value) < 256:
mt_value = hf_value

# Compare values (with tolerance for floats)
is_match = False
if isinstance(hf_value, float) or isinstance(mt_value, float):
Expand Down
Loading