Commit Graph

4 Commits

Author SHA1 Message Date
codingma
9c0b4b3835 [v1][feature] add dpo trainer (#10544)
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-26 15:32:10 +08:00
jiaqiw09
28a6ea1cdc [v1] add deepspeed zero3 trigger for low memory usage weight loading (#10300) 2026-04-21 14:09:52 +08:00
jiaqiw09
df2e6edb7e [v1] add init on rank0 for fsdp2 (#10264) 2026-03-27 14:54:03 +08:00
jiaqiw09
ab073f4c13 [v1] add LoRA/Freeze support and merge workflow (#10157) 2026-02-12 13:02:09 +08:00