Skip to content

Pull requests: huggingface/nanotron

New pull request New
49 Open 192 Closed
49 Open 192 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

DRAFT: Add per domain logging and improve validation mechanism
#350 opened Apr 15, 2025 by paultltc • Draft
6 tasks
fix init and init scaling factor
#349 opened Apr 14, 2025 by NouamaneTazi Loading…
6 tasks
[Feature] Implement CUDA event-based timing for improved GPU performa…
#346 opened Apr 10, 2025 by grewalsk Loading…
5 of 6 tasks
1
5
quicks
#338 opened Apr 4, 2025 by NouamaneTazi • Draft
6 tasks
calcuate mean token accuracy metric while training
#337 opened Apr 4, 2025 by kashif Loading…
[WIP] Add multilingual evals
#336 opened Apr 2, 2025 by anton-l Loading…
6 tasks
Logging outlier batch
#332 opened Apr 1, 2025 by eliebak • Draft
add sanity check to see the batch diversity
#331 opened Apr 1, 2025 by eliebak • Draft
Ademamix
#300 opened Mar 23, 2025 by eliebak • Draft
6 tasks
Muon
#298 opened Mar 23, 2025 by eliebak • Draft
6 tasks
[Feature] Hide 75% of the communication in tensor parallelism using DoMiNo
#292 opened Mar 10, 2025 by xrsrke Loading…
19
[WIP] Distillation
#290 opened Mar 6, 2025 by Stillerman Loading…
2 of 14 tasks
Fix unpacking issue caused by newer Flash Attention
#289 opened Mar 5, 2025 by Stillerman Loading…
3 of 6 tasks
1
2
[Feature] Over 99% communication overlap in Tensor Parallelism using Domino
#286 opened Mar 1, 2025 by hwchen2017 Loading…
4
Recommend the use of Spack on supercomputers
#282 opened Feb 19, 2025 by thomas-bouvier Loading…
Add context manager to increase NCCL timeout temporarily
#280 opened Feb 18, 2025 by thomas-bouvier Loading…
Add MLA
#278 opened Feb 5, 2025 by zzhhjjj Loading…
30
Add nanotron performance
#274 opened Jan 23, 2025 by xrsrke Loading…
[CICD] Add a timeout for unit tests and measure their execution time
#270 opened Jan 17, 2025 by xrsrke Loading…
fp8
#266 opened Dec 18, 2024 by xrsrke Loading…
[Feature] Support resume ZeRO1 in a new data parallelism size
#263 opened Dec 10, 2024 by xrsrke Loading…
Fix wrong initialization of lr scheduler
#256 opened Nov 29, 2024 by kylematoba Loading…
2
[NEW] Llama3.2 weight converters 🦙
#255 opened Nov 28, 2024 by TJ-Solergibert Loading…
6 tasks
4
Add shuffling in Nanotron for subsequent epochs when data is repeated
#247 opened Nov 24, 2024 by Lauler Loading…
3
Fix initial_lr when resuming training
#243 opened Nov 17, 2024 by Lauler Loading…
6
ProTip! Updated in the last three days: updated:>2025-04-12.