Skip to content

Pull requests: InternLM/xtuner

New pull request New
77 Open 773 Closed
77 Open 773 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

fix rl_trainer train_metrics global_step and add avg train_metrics
#1465 opened Jan 30, 2026 by YanhuiDua Loading…
4
Optimize rollout-related ut execution time
#1463 opened Jan 29, 2026 by YanhuiDua Loading…
[Fix] Prevent FSDPModule from auto-converting position_embedding to bf16 in DecoderLayer
#1462 opened Jan 29, 2026 by HAOCHENYE Loading…
[Enhance] Simplify the pipeline of training with hf model
#1461 opened Jan 28, 2026 by HAOCHENYE Loading…
1
[Feature] Provide naive fully shard in BaseModel`
#1460 opened Jan 28, 2026 by HAOCHENYE Loading…
[Fix] Fix bug of chunk_loss when backward with the froze lm_head
#1458 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] update toy tokenizer
#1457 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] Replace annotation of model_cfg with XTunerBaseModelConfig
#1456 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] Add data interface for sequence context
#1455 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] Support using custom collator in DataloaderConfig
#1454 opened Jan 28, 2026 by HAOCHENYE Loading…
[Refactor] Extract common text context and labels building logic to reduce code duplication
#1453 opened Jan 28, 2026 by HAOCHENYE Loading…
[Fix] Fix bug of clean_param_name
#1452 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] Add copy interface for SequenceContext
#1451 opened Jan 28, 2026 by HAOCHENYE Loading…
[Enhance] Remove float8_handler in fully_shared
#1450 opened Jan 28, 2026 by HAOCHENYE Loading…
[Refactor] Replace HF key remapping hack with hf_key_mapping config
#1449 opened Jan 28, 2026 by HAOCHENYE Loading…
qwen3vl support get_model_config_from_hf
#1446 opened Jan 27, 2026 by hhaAndroid Loading…
[Fix] Fix TypedDict import error for Python < 3.12
#1441 opened Jan 23, 2026 by nil0x9 Loading…
[Feat] Support RLDataPacker with several packing strategy
#1438 opened Jan 22, 2026 by YanhuiDua Loading…
12
[Enhance] add prefetch_factor to DataLoaderConfig
#1435 opened Jan 21, 2026 by nil0x9 Loading…
[Don't Merge] add async code
#1432 opened Jan 20, 2026 by hhaAndroid Loading…
CI: use local action and fix spell error
#1429 opened Jan 16, 2026 by kkscilife Loading…
[Feat] support debug_train mode in RL Trainer
#1423 opened Jan 14, 2026 by YanhuiDua Loading…
8
Add ts model
#1420 opened Jan 9, 2026 by hhaAndroid Loading…
[Bug] Applying torch.ops.aten.view.default to a WeightWithDynamicTilewiseFloat8CastTensor will not affect its precomputed_scale and precomputed_w attributes
#1417 opened Jan 8, 2026 by HIT-cwh Loading…
Refactor foreach allgather
#1416 opened Jan 7, 2026 by CyCle1024 Loading…
ProTip! Adding no:label will show everything without a label.