Commits

Branch selector

 develop

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 7, 2025

• BugFix: qwen pp cannot send bool when using fuse loss (#11156)

  Show description for 6ac0402
  Jason233333

  authored
  6ac0402

  Copy full SHA for 6ac0402

• fix zcc ema GPU alloc bug (#11165)

  sneaxiy

  authored
  6ede865

  Copy full SHA for 6ede865

Commits on Nov 5, 2025

• TARE提交 (#11151)

  Show description for 29f97bb
  

  PatriciaPulec

  and

  zhaosiyu-alt

  authored
  29f97bb

  Copy full SHA for 29f97bb

Commits on Nov 3, 2025

• Update run_finetune.py (#11159)

  a31413510

  authored
  10014d8

  Copy full SHA for 10014d8

Commits on Oct 30, 2025

• [codestyle] Modular convert tools (#11129)

  Show description for e73341e
  hsz06

  authored
  e73341e

  Copy full SHA for e73341e

Commits on Oct 27, 2025

• LoRA: add lora target modules support for qwen2 and qwen3 dense model when using fused keys (#11153)

  aiyinyuedejustin

  authored
  1bcf877

  Copy full SHA for 1bcf877

Commits on Oct 24, 2025

• BugFix: qwen model sequence parallel can not get batch size (#11147)

  Show description for c621f0c
  Jason233333

  authored
  c621f0c

  Copy full SHA for c621f0c

Commits on Oct 23, 2025

• paddleformers PR check (#11150)

  Show description for 3f7eafc
  swgu98

  authored
  3f7eafc

  Copy full SHA for 3f7eafc

Commits on Oct 22, 2025

• commit dislora (#11122)

  Show description for 51787ba
  Pioneer-wxh

  authored
  51787ba

  Copy full SHA for 51787ba

• BugFix: Support for Qwen Model PipelineParallel Training in RLInfra (#11143)

  Show description for a4d16fa
  Jason233333

  authored
  a4d16fa

  Copy full SHA for a4d16fa

Commits on Oct 21, 2025

• [AutoParallel]update code (#11145)

  Show description for 7d30769
  xuxinyi389

  authored
  7d30769

  Copy full SHA for 7d30769

Commits on Oct 16, 2025

• Upgrad swandlab code (#11105)

  Show description for 2f78db6
  JunnYu

  authored
  2f78db6

  Copy full SHA for 2f78db6

Commits on Oct 11, 2025

• feat(model): add support for Qwen model in RL PipelineParallel (#11135)

  Jason233333

  authored
  5dd2855

  Copy full SHA for 5dd2855

Commits on Sep 30, 2025

• [Auto-Paralllel] fix intermediate api use1 (#11125)

  Show description for 119ed11
  Xing-lil

  authored
  119ed11

  Copy full SHA for 119ed11

Commits on Sep 29, 2025

• [Embedding] Fix readme and env for contrastive_training (#11108)

  Show description for 3c51eda
  tianyumyum

  authored
  3c51eda

  Copy full SHA for 3c51eda

Commits on Sep 25, 2025

• [CI]Update loss (#11115)

  Show description for 937b766
  Liujie0926

  authored
  937b766

  Copy full SHA for 937b766

• Pr adapt flex checkpoint (#11065)

  Show description for 3432f69
  zty-king

  authored
  3432f69

  Copy full SHA for 3432f69

Commits on Sep 24, 2025

• [CI]Update loss for Paddle#75367 (#11106)

  Show description for 586f5fb
  Liujie0926

  authored
  586f5fb

  Copy full SHA for 586f5fb

Commits on Sep 23, 2025

• [Distributed] refine training_pipeline_step to avoid memory leaks (#11103)

  zhangyuqin1998

  authored
  af07de0

  Copy full SHA for af07de0

• fix approval bypass (#11104)

  Show description for b30ad6a
  swgu98

  authored
  b30ad6a

  Copy full SHA for b30ad6a

Commits on Sep 22, 2025

• fix typos (#11041)

  co63oc

  authored
  615a7c9

  Copy full SHA for 615a7c9

• [AutoParallel] GradientClipByGlobalNorm: Add align mode branch in autotrainer (#10960)

  Show description for 03bc1a0
  zty-king

  authored
  03bc1a0

  Copy full SHA for 03bc1a0

Commits on Sep 15, 2025

• [CI] Update loss_check (#11090)

  Show description for b6f214e
  Liujie0926

  authored
  b6f214e

  Copy full SHA for b6f214e

• Add chunk offload optimizer (#11084)

  Show description for e28c816
  Wennie396

  authored
  e28c816

  Copy full SHA for e28c816

Commits on Sep 11, 2025

• [Embedding] Add qwen3 and update readme (#11034)

  Show description for efd7d26
  Li-Z-Q

  authored
  efd7d26

  Copy full SHA for efd7d26

Commits on Sep 8, 2025

• Make sharding_first by default (#11059)

  Show description for 148206e
  sneaxiy

  authored
  148206e

  Copy full SHA for 148206e

• add bypass (#11077)

  Show description for 462597f
  swgu98

  authored
  462597f

  Copy full SHA for 462597f

• [fea] support dp-moe for zcc and global_expert_id (#11050)

  Show description for 85295b6
  

  bo-ke

  and

  Meiyim

  authored
  85295b6

  Copy full SHA for 85295b6

• [Cherry-pick]Support to load sharded EMA checkpoint (#11075)

  Show description for 509f005
  sneaxiy

  authored
  509f005

  Copy full SHA for 509f005

Commits on Sep 6, 2025

• [cherry-pick] add zcc_ema_loss_threshold args to avoid merging models with loss spike (#11079)

  sjy1203

  authored
  20a584f

  Copy full SHA for 20a584f

Commits on Sep 5, 2025

• Cherry-pick useful PRs from fleety (#11051)

  Show description for 68645cc
  

  SylarTiaNII

  and

  aiyinyuedejustin

  authored
  68645cc

  Copy full SHA for 68645cc

• [PEFT]：nola (#10747)

  Show description for e050fd8
  weiweisss

  authored
  e050fd8

  Copy full SHA for e050fd8

Commits on Sep 4, 2025

• LoriKiT (#9776)

  Show description for e298ba4
  

  Liebele

  and

  bruicecode

  authored
  e298ba4

  Copy full SHA for e298ba4

• use key-value to init dataloader (#11058)

  Waynezee

  authored
  19a585b

  Copy full SHA for 19a585b

• Unified FuseLoss (#11057)

  KB-Ding

  authored
  13790da

  Copy full SHA for 13790da

Pagination

Previous

Next