transformers
b6f9463e - Gemma4 resizing per layer inputs (#45324)

Commit
24 days ago
Gemma4 resizing per layer inputs (#45324) * gemma4 * mask out per-layer input correctly! * revert gemma3n and add t5gemma * fix repo * the test * fix some tests * gemma3n also has a causal model * fix repo * Apply repo consistency fixes * update attr even if weights are tied --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Author
Parents
Loading