Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
263 Open 4,635 Closed
263 Open 4,635 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

[NNX] Remove remaining Linen module code and dead references
#5203 opened Sep 11, 2026 by ecnal-cienet Collaborator • Draft
4 tasks done
6
fix(vllm_decode): optimize hbm_utilization and ici_tensor parallelism for Gemma4-26b
#5202 opened Sep 11, 2026 by chiajunglien Collaborator Loading…
4 tasks done
3
Optimize MaxText cross-entropy loss memory, vocabulary tiling, and PyTree engine caching:
#5201 opened Sep 11, 2026 by copybara-service Bot Loading…
Fix EOS token detection and Qwen vision model routing in multimodal_eval
#5200 opened Sep 11, 2026 by subawocit Collaborator • Draft
4 tasks done
4
Packing adapter segment ids
#5198 opened Sep 11, 2026 by A9isha Collaborator Loading…
4 tasks done
9
PR #4360: [NNX] Delete Linen 4/5: remove the Linen decoder/attention layers and *_as_linen model wrappers
#5197 opened Sep 11, 2026 by copybara-service Bot Loading…
4 tasks done
1
Migrate MaxText Docker images to Artifact Registry
#5196 opened Sep 11, 2026 by SurbhiJainUSC Collaborator • Draft
4 tasks done
Extend MoE dropless retry mechanism to eval step pull ready
#5194 opened Sep 10, 2026 by Shuwen-Fang Collaborator Loading…
4 tasks done
1
Integrate sharded Muon into MaxText.
#5193 opened Sep 10, 2026 by copybara-service Bot Loading…
1
Add Cosmos 3 Core Attention Block to MaxText.
#5192 opened Sep 10, 2026 by copybara-service Bot Loading…
Optimize JAX flash attention for long sequences with a specialized path.
#5191 opened Sep 10, 2026 by copybara-service Bot Loading…
1
[MaxText] Eliminate XLA backward adjoint interior padding in reorder_sequence
#5188 opened Sep 10, 2026 by copybara-service Bot Loading…
1
Docs migration from XPK to CTK(4/5)
#5187 opened Sep 10, 2026 by KevinYenky • Draft
4 tasks done
4
Docs migration from XPK to CTK(5/5)
#5184 opened Sep 10, 2026 by KevinYenky • Draft
4 tasks done
3
Docs migration from XPK to CTK(3/5)
#5183 opened Sep 10, 2026 by KevinYenky • Draft
4 tasks done
6
Onboard Qwen3.5 397b fp8
#5182 opened Sep 9, 2026 by Shuwen-Fang Collaborator Loading…
4 tasks done
13
Fix: Numeric divergence between the mHC layer and its Pallas kernel
#5179 opened Sep 9, 2026 by denis-mil Loading…
4 tasks done
Docs migration from XPK to CTK(1/5)
#5177 opened Sep 9, 2026 by KevinYenky • Draft
4 tasks done
8
Docs migration from XPK to CTK(2/5)
#5176 opened Sep 9, 2026 by KevinYenky • Draft
4 tasks done
9
Docs update from xpk to ctk
#5175 opened Sep 9, 2026 by KevinYenky • Draft
4 tasks
8
Reverts f7f720404a0c6cd47133419edd1a8ff4c5effa3c
#5174 opened Sep 9, 2026 by copybara-service Bot Loading…
1
Onboard the Gemma 4 decoders to explicit sharding
#5162 opened Sep 8, 2026 by NuojCheng Collaborator • Draft
4 tasks done
1
Share one RNG stream across the bridged NVFP4 quantizers
#5160 opened Sep 8, 2026 by ecnal-cienet Collaborator • Draft
4 tasks done
3
Share one CUSTOM_REMAT_TENSORS list between MaxTextConfig and RLConfig
#5159 opened Sep 8, 2026 by lokic233 Loading…
3
Train & Eval on entire dataset when pdbs < 1
#5158 opened Sep 7, 2026 by muskansh-google Contributor • Draft
3
ProTip! Exclude everything labeled bug with -label:bug.