transformers

6082a555 - [Model] Refactor modernbert with the attention interface (#43030)

[Model] Refactor modernbert with the attention interface (#43030) * Refactor FA interface * Simplify code * Update * Simplify the code * Remove comments * Fix UT * Fix _attn_implementation UT * Fix TimmWrapperForImageClassification sdpa dispatch UT * Remove unused parameter * Revise based on comments * Merge main branch * Fix CI error * Fix UT * Refine code based comments * check_config_attributes * Fix UT errors * Refine code * Revert test_modeling_common.py * Revert file * Revert file * Fix rotary(.float()) * Fix silding window errors * Update modeling files * Fix * Fix flash sdpa eager * style * Merge main * Fix CI error * Update * Update the doc * Update code based comments * make fix-repo * Fix CI * check_config_attributes * fix * refactor tests a bit more + fa integration test * style * Fix UTs * resolve conflicts * Update src/transformers/models/modernbert/configuration_modernbert.py Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com> * Revised based on comments. --------- Co-authored-by: Anton Vlasjuk <73884904+vasqu@users.noreply.github.com> Co-authored-by: vasqu <antonprogamer@gmail.com> Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com>