Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
224 Open 4,273 Closed
224 Open 4,273 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 JAX compilation cache clear fixture to prevent XLA memory accumulation gemini-review
#4784 opened Aug 7, 2026 by SurbhiJainUSC Collaborator Loading…
4 tasks done
6
Fix DeepSeek4 shape mismatch in mhc.py and moe.py
#4783 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
4
Fix DeepSeek4HyperHead instantiation and usage in NNXDecoder
#4782 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
4
Fix shape mismatch and add DeepSeek4HyperHead in mhc.py
#4780 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
5
qwen3 muon fork
#4779 opened Aug 7, 2026 by Shuwen-Fang Collaborator Loading…
4 tasks
5
Fix DeepSeek4HyperHead compilation error
#4778 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
4
Fix DeepSeek4HyperHead instantiation and usage in NNXDecoder
#4777 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
3
[maxtext] Add block-diffusion pre-training gemini-review
#4776 opened Aug 7, 2026 by ethannnnnn Collaborator Loading…
4 tasks done
4
Fix DeepSeek4HyperHead instantiation and usage in NNXDecoder
#4775 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
4
Fix DeepSeek4HyperHead instantiation error by using mhc_reduce
#4773 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
1
Fix DeepSeek4HyperHead missing from maxtext.layers.mhc
#4772 opened Aug 7, 2026 by olufiyin19 Collaborator Loading…
3
Fix Qwen 3.5 35B RL gibberish output issue and show end-to-end RL convergence
#4770 opened Aug 7, 2026 by AntonyMei Collaborator Loading…
16
Support Ring Attention with DeepSeek DSA Sparse Indexer gemini-review
#4767 opened Aug 6, 2026 by zcjhao Collaborator Loading…
4 tasks done
14
Support gemma4 decoder under explicit sharding mode
#4766 opened Aug 6, 2026 by csgoogle Contributor Loading…
1 of 2 tasks
4
Add compare_training_engine.py test suite comparing standalone pre_train.train_step against MaxTextTrainingEngine.
#4762 opened Aug 6, 2026 by copybara-service Bot Loading…
@A9isha
1
Moe dynamic redundancy
#4760 opened Aug 6, 2026 by snehalv2002 Collaborator • Draft
4 tasks
7
[DiLoCo] Streaming diloco with vmap (SPMD Version) enhancement New feature or request gemini-review
#4759 opened Aug 6, 2026 by Dr-Left Collaborator Loading…
4 tasks done
@khatwanimohit
30
Add Grain support for multimodal SFT data processing pipeline
#4754 opened Aug 6, 2026 by snehalv2002 Collaborator Loading…
4 tasks done
9
Enabled auxilillary loss free load balancing and sequence wise load balancing for DSV4 [Replacement for PR-https://github.com/AI-Hypercomputer/maxtext/pull/4233] pull ready
#4753 opened Aug 6, 2026 by dipakg-lang Collaborator Loading…
4 tasks done
6
Download tokenizers from GCS in unit tests instead of HF pull ready
#4752 opened Aug 5, 2026 by igorts-git Collaborator Loading…
4 tasks done
10
Dsv4 pm
#4751 opened Aug 5, 2026 by snehalv2002 Collaborator • Draft
4 tasks
5
fix(pylint): ignore datasets module to resolve E0611 no-name-in-module errors
#4750 opened Aug 5, 2026 by NuojCheng Collaborator Loading…
3
Add README to maxtext.trainers gemini-review
#4749 opened Aug 5, 2026 by melissawm Collaborator Loading…
4 tasks done
3
[No merge] Draft test pull ready
#4747 opened Aug 5, 2026 by NuojCheng Collaborator • Draft
4 tasks
Decouple vision & decoder block routing in multimodal processor
#4746 opened Aug 5, 2026 by subawocit Collaborator Loading…
4 tasks done
8
ProTip! Exclude everything labeled bug with -label:bug.