Yan Bai dd8864f9ee [megatron] feat: script of qwen3vl 235b (#3799) ...

an example script

2025-10-17 16:46:45 +08:00

..

data_preprocess

[tool] feat: support load local datasets when preparing datasets (#3621)

2025-09-26 11:42:53 +08:00

generation

[ci] refactor: setup testing guidance (#1958)

2025-06-12 06:16:58 -07:00

gmpo_trainer

[recipe] fix: update readme for gmpo-trainer (#3764)

2025-10-15 10:24:24 +08:00

gpg_trainer

[BREAKING][megatron] refactor: activation checkpointing APIs (#2651)

2025-07-22 10:24:28 +08:00

grpo_trainer

[megatron] feat: script of qwen3vl 235b (#3799)

2025-10-17 16:46:45 +08:00

ppo_trainer

[doc] fix: actor_rollout_ref.critic is not correct (#3778)

2025-10-16 11:12:45 +08:00

ray

[ci] refactor: reduce ruff line-length from 300 to 120 (#2287)

2025-07-01 09:54:40 +08:00

reinforce_plus_plus_trainer

set use_kl_in_reward=True in reinforce_plus_plus (#2580)

2025-07-17 12:10:54 +08:00

remax_trainer

[doc] fix: quickstart example can't work on zsh (#2509)

2025-07-14 13:26:32 +08:00

rloo_trainer

[doc] fix: quickstart example can't work on zsh (#2509)

2025-07-14 13:26:32 +08:00

rollout_importance_sampling

[rollout] refactor: rename "clip" mode back to "mask" mode (#3750)

2025-10-13 11:06:36 -07:00

sft

[model] feat: support ByteDance Seed-OSS 36B model (#3347)

2025-09-04 22:41:58 +08:00

sglang_multiturn

[sglang, recipe] feat: add SGLang as rollout engine for one-step-off-policy (#3531)

2025-10-14 10:48:29 +08:00

skypilot

[deployment, doc] feat: Add SkyPilot integration examples (#3333)

2025-09-04 16:56:00 +08:00

slurm

[doc] fix: quickstart example can't work on zsh (#2509)

2025-07-14 13:26:32 +08:00

split_placement

[trainer] feat: ReMax support using reward model for baseline (#3780)

2025-10-17 12:07:05 +08:00

tuning

[recipe] feat: CollabLLM integration for multiturn training (#3574)

2025-09-25 09:53:39 +08:00

tutorial/agent_loop_get_started

[doc] chore: add agent loop get started tutorial (#3790)

2025-10-17 08:30:10 +08:00