Skip to content

Bumping flash attention version to 2.6.3 and adding option for softcap in attention and lm_head logits. #4452

Bumping flash attention version to 2.6.3 and adding option for softcap in attention and lm_head logits.

Bumping flash attention version to 2.6.3 and adding option for softcap in attention and lm_head logits. #4452

smoketest (3.9)

succeeded Sep 22, 2024 in 1m 32s