Skip to content

Pull requests: PaddlePaddle/PaddleNLP

New pull request New
Clear current search query, filters, and sorts
24 Open 4,055 Closed
24 Open 4,055 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[Auto-Paralllel] fix intermediate api use1
#11125 by Xing-lil was merged Sep 30, 2025 Loading…
2 tasks
3
Fix empty usage error of aux-loss-free
#11116 by lshpku was merged Sep 26, 2025 Loading…
fix readme and env Beijing Innovation Consortium
#11108 by tianyumyum was merged Sep 29, 2025 Loading…
2
fix approval bypass
#11104 by swgu98 was merged Sep 23, 2025 Loading…
2 tasks
1
[Distributed] refine training_pipeline_step to avoid memory leaks
#11103 by zhangyuqin1998 was merged Sep 23, 2025 Loading…
4
update post norm rc
#11097 by chen2016013 was merged Sep 23, 2025 Loading…
2 tasks
1
Fix FusedRMSLinear backward compute
#11095 by lshpku was merged Sep 23, 2025 Loading…
1
add chunk offload optimizer
#11084 by Wennie396 was merged Sep 15, 2025 Loading…
2
[cherry-pick] add zcc_ema_loss_threshold args to avoid merging models… contributor
#11079 by sjy1203 was merged Sep 6, 2025 Loading…
2 tasks
@gongel
3
add bypass skip-ci: distribute-a100 skip-ci: distribute-v100
#11077 by swgu98 was merged Sep 8, 2025 Loading…
2 tasks
1
Support to load sharded EMA checkpoint without loading non-EMA checkpoints
#11076 by sneaxiy was merged Sep 6, 2025 Loading…
2 tasks
2
[Cherry-pick]Support to load sharded EMA checkpoint
#11075 by sneaxiy was merged Sep 8, 2025 Loading…
2 tasks
3
Support to load sharded EMA checkpoint
#11073 by sneaxiy was merged Sep 6, 2025 Loading…
2 tasks
2
use key-value to init dataloader
#11067 by Waynezee was merged Sep 4, 2025 Loading…
2 tasks
3
token_dispatcher support expert_num 64
#11066 by blacksheep-Aristotle was merged Sep 4, 2025 Loading…
2 tasks
2
Pr adapt flex checkpoint contributor
#11065 by zty-king was merged Sep 25, 2025 Loading…
@KB-Ding
28
fix custom op compile bug
#11064 by risemeup1 was merged Sep 4, 2025 Loading…
2 tasks
3
Fix post_process_backward calc stream wait
#11063 by lshpku was merged Sep 4, 2025 Loading…
1
Make sharding_first by default skip-ci: distribute-a100 skip-ci: distribute-v100
#11059 by sneaxiy was merged Sep 8, 2025 Loading…
2 tasks
2
Unified FuseHeadLoss
#11057 by KB-Ding was merged Sep 4, 2025 Loading…
1
Adapting Qwen3 FuseLoss
#11056 by KB-Ding was merged Sep 4, 2025 Loading…
1
[Cherry-pick] Cherry-pick hybrid-expert parallel related PRs from fleety
#11052 by pkuzyc was merged Sep 3, 2025 Loading…
2
Cherry-pick useful PRs from fleety
#11051 by SylarTiaNII was merged Sep 5, 2025 Loading…
2
[fea] support dp-moe for zcc and global_expert_id contributor
#11050 by bo-ke was merged Sep 8, 2025 Loading…
2 tasks
@DesmonDay
2
Cherry-pick PRs from incubate/paddlenlp_fleety
#11049 by sneaxiy was merged Sep 3, 2025 Loading…
2 tasks
2
ProTip! Add no:assignee to see everything that’s not assigned.