Commit Graph

4 Commits

Author SHA1 Message Date
Jaret Burkett
524bd2edfc Make flash attn optional. Handle larger batch sizes. 2025-04-14 14:34:46 +00:00
Jaret Burkett
3a5ea2c742 Remove some moe stuff for finetuning. Drastically reduces vram usage 2025-04-14 00:57:34 +00:00
Jaret Burkett
f80cf99f40 Hidream is training, but has a memory leak 2025-04-13 23:28:18 +00:00
Jaret Burkett
594e166ca3 Initial support for hidream. Still a WIP 2025-04-13 13:50:11 -06:00