Skip to content

Pull requests: NVIDIA-NeMo/Megatron-Bridge

New pull request New
59 Open 608 Closed
59 Open 608 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

Gemma3 Provider & Bridge
#867 opened Oct 3, 2025 by suiyoubi Loading…
3
Support chat templatization and tool calling in SFT datasets
#864 opened Oct 3, 2025 by ananthsub • Draft
2
@ananthsub
3
allow specifying kwargs for hf tokenizer in the config
#863 opened Oct 3, 2025 by ananthsub Loading…
2
Add 0.1.0rc3 changelog r0.1.0
#862 opened Oct 3, 2025 by chtruong814 • Draft
1
Add functional test for finetuning with sequence packing testing Request for or addition to test suite
#861 opened Oct 3, 2025 by ananthsub • Draft
2
Optimize disk management
#858 opened Oct 2, 2025 by pablo-garay Loading…
3
[sync] Fix bug in param_norm computation where some ranks might call collective and some might not
#852 opened Oct 2, 2025 by ananthsub Loading…
2
[nemo run] Allow plugins to customize script args
#848 opened Oct 1, 2025 by ananthsub Loading…
1
2
support profiling with nvtx ranges in megatron core
#847 opened Oct 1, 2025 by ananthsub Loading…
2
Support scaling runs for DeepSeek-V3
#834 opened Oct 1, 2025 by scsudhakaran Loading…
1
dsv3 blackwell config update performance r0.1.0
#832 opened Oct 1, 2025 by malay-nagda Loading…
1
Gdeng/malay/dsv3 2509
#831 opened Oct 1, 2025 by gdengk • Draft
1
Qwen 25vl training pipeline
#830 opened Oct 1, 2025 by yaoyu-33 Loading…
13
Cherry pick qwen2.5 VL nemo-rl changes
#828 opened Sep 30, 2025 by yfw Loading…
1
Fix DP size calculation in config.py
#819 opened Sep 29, 2025 by yaoyu-33 Loading…
1
[docs] Update config overview with more motivation behind the design docs-only With great power comes great responsibility. documentation Improvements or additions to documentation
#808 opened Sep 27, 2025 by ananthsub Loading…
8
Improve typehints for forward_step_func and add tests for functor inputs
#806 opened Sep 27, 2025 by ananthsub Loading…
8
extra validation for in process restart to assert process group not initialized before training
#805 opened Sep 27, 2025 by ananthsub Loading…
@ananthsub
2
Support skipping iterations with dummy step
#802 opened Sep 27, 2025 by ananthsub Loading…
@ananthsub
3
Support sample-based training
#801 opened Sep 27, 2025 by ananthsub Loading…
1
@ananthsub
3
[sync] meta device and fsdp init
#800 opened Sep 27, 2025 by ananthsub • Draft
2
Conditionally destroy process group at end of train
#795 opened Sep 26, 2025 by ananthsub Loading…
1
4
rename seq_length to sequence_length
#782 opened Sep 26, 2025 by dimapihtar Loading…
3
add dclm tutorial
#781 opened Sep 26, 2025 by dimapihtar Loading…
10
[Draft] Qwen3-VL Support
#776 opened Sep 26, 2025 by yaoyu-33 Loading… 25.11
1
ProTip! Follow long discussions with comments:>50.