Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Jul 9, 2025

• [cfg] refactor: make the rollout & ref configs more modular (#2410)

  Show description for cccc2ef
  eric-haibin-lin

  authored
  cccc2efView commit details

  Copy full SHA for cccc2ef

  Browse repository at this point

• [sglang] fix: Bug in megatron+sglang TP16 update_weights. (#2336)

  Show description for ad33564
  SuperCB

  authored
  ad33564View commit details

  Copy full SHA for ad33564

  Browse repository at this point

• [data] refactor: move sampler api to experimental (#2381)

  eric-haibin-lin

  authored
  4def91dView commit details

  Copy full SHA for 4def91d

  Browse repository at this point

• [rollout] fix: huggingface model config max_position_embeddings assertion for model with extended context length (#737)

  Show description for 004da73
  

  3 peopleauthored
  004da73View commit details

  Copy full SHA for 004da73

  Browse repository at this point

Commits on Jul 8, 2025

• [ci] feat: add docstring checker script and comprehensive docstrings (#2378)

  Show description for a4033af
  

  eric-haibin-lin

  and

  devin-ai-integration[bot]

  authored
  a4033afView commit details

  Copy full SHA for a4033af

  Browse repository at this point

• [ci] fix: forbid ci on forks (#2412)

  Show description for 588f972
  plutoZZZZ

  authored
  588f972View commit details

  Copy full SHA for 588f972

  Browse repository at this point

• [misc] feat: trace rollout generation and tool calls using weave (#2345)

  Show description for ec4433c
  chenhaiq

  authored
  ec4433cView commit details

  Copy full SHA for ec4433c

  Browse repository at this point

• [sglang] fix: Import Error in the latest sglang (#2275)

  Show description for 578501e
  yyDing1

  authored
  578501eView commit details

  Copy full SHA for 578501e

  Browse repository at this point

Commits on Jul 7, 2025

• [sglang] fix: only wake up weights on infer_tp 0 (#2403)

  zhaochenyang20

  authored
  ee65422View commit details

  Copy full SHA for ee65422

  Browse repository at this point

• [cfg] refactor: make actor config more modular (#2379)

  eric-haibin-lin

  authored
  3f929afView commit details

  Copy full SHA for 3f929af

  Browse repository at this point

• [tool] fix: Add MCP usage documentation (#2261)

  AlecHenx

  authored
  1e7c545View commit details

  Copy full SHA for 1e7c545

  Browse repository at this point

• [ci] feat: use action (#2393)

  Show description for cb3dcc6
  plutoZZZZ

  authored
  cb3dcc6View commit details

  Copy full SHA for cb3dcc6

  Browse repository at this point

• [sglang, rollout, doc] fix: update sglang rollout generate doc (#2385)

  Show description for 26e26d1
  Tavish9

  authored
  26e26d1View commit details

  Copy full SHA for 26e26d1

  Browse repository at this point

• [misc] refactor: replace pkg_resources with importlib.metadata (#2392)

  Show description for 26d3a03
  askender

  authored
  26d3a03View commit details

  Copy full SHA for 26d3a03

  Browse repository at this point

• [BREAKING][rollout] feat: repeat DataProto when n>1 in driver instead of rollout workers (#2324)

  Show description for fc35956
  

  wuxibin89

  and

  zhaochenyang20

  authored
  fc35956View commit details

  Copy full SHA for fc35956

  Browse repository at this point

• [rollout] fix: sglang async fail with Multi-stage Awake feature (#2365)

  Show description for 4c37c97
  chenhaiq

  authored
  4c37c97View commit details

  Copy full SHA for 4c37c97

  Browse repository at this point

Commits on Jul 6, 2025

• [trainer] fix: Use safe masked mean/sum to handle NaN values outside the mask (#2377)

  Show description for 5cbad83
  Yangruipis

  authored
  5cbad83View commit details

  Copy full SHA for 5cbad83

  Browse repository at this point

• [sglang, rollout] refactor: use torch.Tensor in async rollout schemas (#2362)

  nanjiangwill

  authored
  891c873View commit details

  Copy full SHA for 891c873

  Browse repository at this point

• [ci] fix: PR title check supports module names with underscore (`training_utils`) (#2383)

  HollowMan6

  authored
  2a01b21View commit details

  Copy full SHA for 2a01b21

  Browse repository at this point

• [doc] feat: add July events (#2382)

  eric-haibin-lin

  authored
  c71fa39View commit details

  Copy full SHA for c71fa39

  Browse repository at this point

Commits on Jul 5, 2025

• [doc] fix: Fix document config.rst (#2369)

  Show description for 281ecd4
  shuyhere

  authored
  281ecd4View commit details

  Copy full SHA for 281ecd4

  Browse repository at this point

• Revert "[misc] fix: invalid escape sequence '\*'" (#2376)

  Show description for e9b38dc
  vermouth1992

  authored
  e9b38dcView commit details

  Copy full SHA for e9b38dc

  Browse repository at this point

• [rollout] fix: fix hf rollout and add single gpu test (#2371)

  eric-haibin-lin

  authored
  cbeb3f4View commit details

  Copy full SHA for cbeb3f4

  Browse repository at this point

• [misc] fix: invalid escape sequence '\*' (#2375)

  Show description for 50ba712
  HollowMan6

  authored
  50ba712View commit details

  Copy full SHA for 50ba712

  Browse repository at this point

• [ray] refactor: Seperate the constants into different file (#2025)

  Show description for 9cc3077
  

  YeonwooSung

  and

  eric-haibin-lin

  authored
  9cc3077View commit details

  Copy full SHA for 9cc3077

  Browse repository at this point

• [doc] fix: add show source option (#2370)

  Show description for 9b0e327
  eric-haibin-lin

  authored
  9b0e327View commit details

  Copy full SHA for 9b0e327

  Browse repository at this point

Commits on Jul 4, 2025

• [tool] feat: Add support for tools that generate multimodal data (#2146)

  Show description for 715724c
  nanjiangwill

  authored
  715724cView commit details

  Copy full SHA for 715724c

  Browse repository at this point

• [cfg] fix: pickleing error in multiprocessing in the reward_fn (#2239)

  Show description for 1b891dc
  

  3 peopleauthored
  1b891dcView commit details

  Copy full SHA for 1b891dc

  Browse repository at this point

• [data] feat: add interface for user-defined curriculum sampler (#2314)

  Show description for dbd4ff1
  

  4 peopleauthored
  dbd4ff1View commit details

  Copy full SHA for dbd4ff1

  Browse repository at this point

• [trainer, cfg] feat: add BaseConfig for all dataclass configs. Introduce dataclass for algorithm related configs (#2147)

  Show description for c936ec7
  

  3 peopleauthored
  c936ec7View commit details

  Copy full SHA for c936ec7

  Browse repository at this point

• [hardware] feat: support ray actor sharing situation on ASCEND NPU (#2341)

  Show description for 5c39b51
  FightingZhen

  authored
  5c39b51View commit details

  Copy full SHA for 5c39b51

  Browse repository at this point

• [trainer] fix: pre-commit broken by #2354 (#2358)

  Show description for 8883b29
  ETOgaosion

  authored
  8883b29View commit details

  Copy full SHA for 8883b29

  Browse repository at this point

• [rollout] fix: #1646 stop words for sglang rollout (#1991)

  Show description for 0d2af47
  

  linxxx3

  and

  zhaochenyang20

  authored
  0d2af47View commit details

  Copy full SHA for 0d2af47

  Browse repository at this point

• [docker] refactor: Migrate images to verlai, support latest flash attention and newer CUDA versions in future (#2085)

  Show description for ebb21b7
  ETOgaosion

  authored
  ebb21b7View commit details

  Copy full SHA for ebb21b7

  Browse repository at this point

• [ckpt] fix: edit esi doc (#2354)

  Show description for a53fb30
  plutoZZZZ

  authored
  a53fb30View commit details

  Copy full SHA for a53fb30

  Browse repository at this point