transformers
83a6c5b5 - Remove cache_position in more models (3) (#44759)

Commit
1 day ago
Remove cache_position in more models (3) (#44759) * start on the mambas * fix mambas * moshi and kyutai * a few more special ones * refactor recurrent gemma * a bit more * zambas * fix mamba * fixes * fix csm * kyutao * fix test * align and simpolify mamba cache * fix mask for recurrent gemma * small oupsi * align falcon_h1 + other review comments
Author
Parents
Loading