cxy c4e09c7cbe [v1] support GDN Ulysses cp (#10727) ...

Co-authored-by: cxy-thinkbook <xuanyuchen@seu.edu.cn>

2026-08-20 18:53:39 +08:00

..

api

[fix] handle empty content list in system message (#10291)

2026-03-18 12:05:49 +08:00

chat

[webui] add seed controls for reproducibility (#10629)

2026-07-07 17:30:09 +08:00

data

[data] pad position_ids on non-FA2 packing path (fixes rotary crash for Gemma-3/4) (#10737)

2026-08-09 16:00:10 +08:00

eval

[misc] update format (#7277)

2025-03-13 02:53:08 +08:00

extras

[model] add MOSS-VL support (#10708)

2026-08-03 18:18:24 +08:00

hparams

[train] Add KTransformers VLM fine-tuning support (#10760)

2026-08-18 19:49:47 +08:00

model

[train] Add KTransformers VLM fine-tuning support (#10760)

2026-08-18 19:49:47 +08:00

third_party

[model] Patch GDN for NPU (#10504)

2026-06-04 16:39:02 +08:00

train

[train] Fix hyper parallel tail accumulation loss scaling (#10705)

2026-07-30 17:29:59 +08:00

v1

[v1] support GDN Ulysses cp (#10727)

2026-08-20 18:53:39 +08:00

webui

[ci] pin ruff version and fix lint errors (#10681)

2026-07-24 16:29:58 +08:00

__init__.py

[trainer] fix pt loss (#7748)

2025-04-17 03:15:35 +08:00

cli.py

[v1] add v1 launcher (#9236)

2025-10-07 22:34:48 +08:00

launcher.py

[train] support megatron-bridge for PT/SFT training (#10645)

2026-07-27 18:45:18 +08:00