Skip to content

Pull requests: NVIDIA-NeMo/Megatron-Bridge

New pull request New
148 Open 1,466 Closed
148 Open 1,466 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

docs: add Kimi K2, GLM-4.5V and fix Qwen3-VL link
#2108 opened Jan 28, 2026 by sbhavani Loading…
2 of 5 tasks
1
Add max_seq_length to vlm_inference.py
#2106 opened Jan 28, 2026 by huvunvidia Loading…
5 tasks
4
Mbridge Deterministic Recipe example
#2103 opened Jan 28, 2026 by suiyoubi Loading…
5 tasks
Documentation improvement for HF <-> Megatron checkpoint conversion
#2102 opened Jan 28, 2026 by OlegSudakov Loading…
5 tasks
3
fix(models): fire language model of qwen3vlmoe by default community-request
#2098 opened Jan 28, 2026 by Yangruipis Loading…
5 tasks
4
kimi k2 recipe intro
#2097 opened Jan 28, 2026 by malay-nagda • Draft
5 tasks
1
📝 Add docstrings to callback
#2094 opened Jan 28, 2026 by coderabbitai bot Loading…
@yaoyu-33
1
[sync] DeepSeek V3 FSDP Fix for Precision-Aware Optimizer
#2093 opened Jan 27, 2026 by ananthsub • Draft
2 of 5 tasks
2
[sync] Update RNG sharding to include EP rank
#2092 opened Jan 27, 2026 by ananthsub • Draft
5 tasks
2
[sync] Save memory using main_param for moe in param_l2_norm
#2091 opened Jan 27, 2026 by ananthsub • Draft
1 of 5 tasks
2
[sync] Various quality-of-life improvements in training loop
#2090 opened Jan 27, 2026 by ananthsub • Draft
2 of 5 tasks
2
[sync] Add support for fake distributed process groups
#2089 opened Jan 27, 2026 by ananthsub • Draft
2 of 5 tasks
2
[sync] Reduce Overhead in Timers
#2088 opened Jan 27, 2026 by ananthsub Loading…
1 of 5 tasks
3
Change exp_name to LLMB convention
#2087 opened Jan 27, 2026 by nv-mollys Loading…
5 tasks
10
Set HF_HUB_OFFLINE=1 by default, enabled with '--hf_token' flag.
#2086 opened Jan 27, 2026 by sudostock Loading…
6
Add container-writable flag.
#2084 opened Jan 27, 2026 by sudostock Loading…
3
DSV3 NVFP4 recipe on GB300
#2076 opened Jan 27, 2026 by dingqingy-nv Loading…
7
ci: Extract and test memory
#2072 opened Jan 26, 2026 by ko3n1g Loading…
5 tasks
5
[misc] fix: Change cached_property to property for _causal_lm_architecture
#2070 opened Jan 26, 2026 by guyueh1 Loading…
5 tasks
6
[build] chore: Upgrade transformers to 5.0
#2068 opened Jan 26, 2026 by yaoyu-33 Loading…
7
Add refactored recipe files for pretrain configs of LLMs
#2067 opened Jan 26, 2026 by athitten Loading…
5 tasks
15
add FIM dataset support
#2066 opened Jan 26, 2026 by dimapihtar Loading…
5 tasks
10
feat: Support callbacks for third-party extensions
#2063 opened Jan 26, 2026 by ananthsub Loading…
3 of 5 tasks
1
11
[misc] refactor: Replace deprecated typing imports (#993) community-request
#2057 opened Jan 24, 2026 by KunalSachdev2005 Loading…
1 of 5 tasks
1
4
[model, refactor] refactor: Centralize provider_bridge config mapping in base class
#2052 opened Jan 23, 2026 by yaoyu-33 Loading…
19 of 21 tasks
19
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.