Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
150 Open 3,396 Closed
150 Open 3,396 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

Add utility to reshard TFDS dataset for large scale training
#3789 opened May 1, 2026 by RissyRan Collaborator Loading…
4 tasks done
Build MaxText docker image from PyPI
#3788 opened May 1, 2026 by SurbhiJainUSC Collaborator • Draft
4 tasks done
1
Parameterize bos_id in inference_microbenchmark.py pull ready
#3786 opened Apr 30, 2026 by A9isha Collaborator Loading…
4 tasks done
1
[WIP] Multimodal quality benchmark
#3784 opened Apr 30, 2026 by hengtaoguo Collaborator • Draft
4 tasks
Exclude RNG state when updating vllm with checkpoint for pre-rl evaluation
#3782 opened Apr 30, 2026 by SurbhiJainUSC Collaborator Loading…
4 tasks done
Feat/Support-Qwix-Quantization-For-NNX-Models
#3781 opened Apr 30, 2026 by hsuan-lun-chiang Collaborator • Draft
4 tasks
Support specifying tokamax gmm tile sizes in MaxText
#3779 opened Apr 29, 2026 by darisoy Collaborator Loading…
4 tasks done
Reorganize pre-training doc
#3778 opened Apr 29, 2026 by melissawm Collaborator Loading…
2 tasks done
9
Add unit tests for evaluate_rl.py
#3776 opened Apr 29, 2026 by SurbhiJainUSC Collaborator Loading…
4 tasks done
Remove local sort after ragged all-to-all
#3774 opened Apr 29, 2026 by copybara-service Bot Loading…
[NNX] NNX migration prep (6/N): NNX-native DPO
#3773 opened Apr 29, 2026 by ecnal-cienet Collaborator • Draft
4 tasks done
FP8 Megablox for batch split pull ready
#3770 opened Apr 29, 2026 by BirdsOfAFthr Collaborator Loading…
4 tasks done
8
generalized LTI
#3769 opened Apr 29, 2026 by vlad-karp Collaborator • Draft
4 tasks
[NNX] NNX migration prep (5/N): correctness fixes and feature enablements
#3766 opened Apr 28, 2026 by ecnal-cienet Collaborator • Draft
4 tasks done
Refactor moe sharding logic into smaller functions pull ready
#3763 opened Apr 28, 2026 by gobbleturk Collaborator Loading…
4 tasks done
4
Add TP support with vLLM FusedMoE kernel
#3762 opened Apr 28, 2026 by NicoGrande Collaborator Loading…
4 tasks done
8
CI/CD UT optimization
#3753 opened Apr 27, 2026 by charlesli640 Collaborator Loading…
4 tasks done
1
[RFC][Docs] Design for JAX-MCTS (Tree of Thoughts) Inference Strategy
#3751 opened Apr 27, 2026 by yipkingster • Draft
3 of 4 tasks
Add dataset type olmo_grain for AI2 OLMo numpy pretrain mixes gemini-review
#3749 opened Apr 26, 2026 by gagika Collaborator Loading…
4 tasks done
7
Internal change
#3743 opened Apr 24, 2026 by copybara-service Bot Loading…
@SurbhiJainUSC
1
Replace hardcoded bucket names with generic ones/env variables
#3738 opened Apr 24, 2026 by melissawm Collaborator Loading…
1 task done
1
Add fused_moe_mlp: fuse wi_0 and wi_1 into one grouped GEMM for MoE FFN1
#3736 opened Apr 23, 2026 by abhinavgoel95 Contributor Loading…
4 tasks done
8
Add support for flexible MTP layer architecture in MaxText.
#3734 opened Apr 23, 2026 by parambole Collaborator • Draft
4 tasks
1
Add some log to debug AddLabel workflow stale Automatically applied to stale PRs.
#3723 opened Apr 22, 2026 by charlesli640 Collaborator • Draft
4 tasks done
1
Replace references to model_creation_utils.create_nnx_model.
#3720 opened Apr 22, 2026 by copybara-service Bot Loading…
4 tasks done
2
ProTip! Updated in the last three days: updated:>2026-04-28.