transformers
f84d85ba - [`FA-2`] Add Flash Attention to `Phi` (#27661)

Commit
2 years ago
[`FA-2`] Add Flash Attention to `Phi` (#27661) * add FA and modify doc file * test_flash_attn_2_generate_padding_right test overwritten * comment * modify persimmon modeling file * added speedup graph * more changes
Author
Parents
Loading