Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
172 Open 3,191 Closed
172 Open 3,191 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

Bump the pip group across 1 directory with 3 updates dependencies Pull requests that update a dependency file python Pull requests that update python code
#3602 opened Apr 8, 2026 by dependabot bot Loading…
1
docs: add LoRA fine-tuning tutorial
#3601 opened Apr 8, 2026 by chiajunglien • Draft
4 tasks
[DO NOT SUBMIT] EP-TP proof of concept (only supports EP + DP)
#3600 opened Apr 8, 2026 by gobbleturk • Draft
4 tasks
1
Remove unnecessary sharding logic for configuring Orbax single replica restore and broadcast.
#3598 opened Apr 7, 2026 by copybara-service bot Loading…
1
Fix Metric Reporting Crash (ZeroDivisionError) using Tunix PerfTracer
#3597 opened Apr 7, 2026 by igorts-git • Draft
4 tasks done
1
Fix explicit sharding for moe models
#3595 opened Apr 7, 2026 by Shuwen-Fang Loading…
4 tasks done
@Shuwen-Fang
1
Enhance pre-commit output in CodeQuality workflow.
#3593 opened Apr 7, 2026 by copybara-service bot Loading…
@kanglant
1
Add Gemma 4 FLOPs & fix sliding window flops computations gemini-review
#3592 opened Apr 7, 2026 by gagika Loading…
4 tasks done
20
Using synthetic dataset_type for more cases
#3591 opened Apr 7, 2026 by charlesli640 Loading…
4 tasks done
3
Fix SequenceDescriptor API for TransformerEngine >= 2.12
#3589 opened Apr 7, 2026 by yeandy • Draft
4 tasks
Fix qwen2 vllm_mapping for tunix pull ready
#3588 opened Apr 7, 2026 by ChingTsai Loading…
4 tasks done
1
Standardize SFT model interface with TunixMaxTextAdapter
#3584 opened Apr 6, 2026 by igorts-git • Draft
4 tasks
1
adding support for None chat templates.
#3582 opened Apr 6, 2026 by NicoGrande Loading…
4 tasks done
1
Fix ZeroDivisionError in SFT during record_train_metrics
#3581 opened Apr 6, 2026 by igorts-git • Draft
4 tasks done
1
Internal test about EP
#3579 opened Apr 6, 2026 by RissyRan • Draft
1
Remove no exp usage from logical rule Part I
#3578 opened Apr 6, 2026 by NuojCheng Loading…
3 of 4 tasks
6
Try to use SLURM_STEP_GPUS for device list if CUDA_VISIBLE_DEVICES is not set
#3577 opened Apr 6, 2026 by gabeweisz Loading…
4 tasks done
1
3
Standardize SFT-Tunix integration and add manual step timing
#3576 opened Apr 6, 2026 by igorts-git • Draft
4 tasks
1
[WIP] Don't review.
#3575 opened Apr 6, 2026 by igorts-git • Draft
4 tasks
1
Adding gather reduce sparse core kernel. pull ready
#3573 opened Apr 4, 2026 by clee1994 Loading…
4 tasks done
Ragged buffer size control
#3569 opened Apr 3, 2026 by copybara-service bot Loading…
@gobbleturk
1
Fix link to checkpointing flags, and flag name
#3566 opened Apr 3, 2026 by copybara-service bot Loading…
@shralex
DeepSeek V3.2 user guide update gemini-review
#3565 opened Apr 3, 2026 by snehalv2002 Loading…
38
Add ragged buffer factor draft Draft PR
#3564 opened Apr 3, 2026 by NuojCheng • Draft
4 tasks
1
Weight structured sparsity in training
#3561 opened Apr 2, 2026 by BirdsOfAFthr • Draft
4 tasks done
ProTip! Filter pull requests by the default branch with base:main.