Skip to content

Pull requests: AI-Hypercomputer/maxtext

New pull request New
189 Open 3,937 Closed
189 Open 3,937 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

[CI] Split TPU and GPU nightly builds to run E2E tests in parallel
#4400 opened Jul 9, 2026 by RexBearIU Collaborator Loading…
Split TPU and GPU nightly image builds to run E2E tests in parallel:
#4398 opened Jul 9, 2026 by chiajunglien Contributor • Draft
4 tasks done
1
Fix ragged all-to-all with ragged buffer factor in DeepSeek-V3.
#4397 opened Jul 9, 2026 by copybara-service Bot Loading…
@RissyRan
Separate RL configurations into a dedicated config class.
#4395 opened Jul 9, 2026 by copybara-service Bot Loading…
@SurbhiJainUSC
Use setup_training_state to fix standalone checkpoint unit tests pull ready
#4394 opened Jul 9, 2026 by xibinliu Collaborator • Draft
4 tasks done
1
Add v6e-8 GSM8K RL convergence benchmark config
#4393 opened Jul 8, 2026 by AntonyMei Collaborator Loading…
4 tasks done
2
Fix RL LR schedule default
#4392 opened Jul 8, 2026 by AntonyMei Collaborator Loading…
4 tasks done
1
[Do not submit] test2 pull ready
#4390 opened Jul 8, 2026 by aireenmei Collaborator • Draft
4 tasks done
Add RL tutorial and run script for GPT-OSS 20B on Multi-Host TPUs
#4389 opened Jul 8, 2026 by susanbao Collaborator Loading…
3 tasks done
1
Getting started with Maxtext with Cluster toolkit doc
#4386 opened Jul 8, 2026 by darisoy Collaborator • Draft
4 tasks done
Fix incorrect sharding for TSP, DP, fused_mlp workloads
#4385 opened Jul 8, 2026 by dinodeep Collaborator Loading…
4 tasks done
1
Use helper functions for CheckpointManager operations.
#4383 opened Jul 8, 2026 by copybara-service Bot Loading…
1
Isolate Orbax v0 emergency checkpointing.
#4382 opened Jul 8, 2026 by copybara-service Bot Loading…
1
Add checkpoint_context.py to map flat orbax v0 flags to Orbax v1 Context.
#4381 opened Jul 8, 2026 by copybara-service Bot Loading…
1
Add Orbax v1 StatefulCheckpointable for MaxText grain iterators.
#4380 opened Jul 8, 2026 by copybara-service Bot Loading…
1
Refactor grain checkpoint handler functionality into grain_utility file.
#4379 opened Jul 8, 2026 by copybara-service Bot Loading…
1
[Gated Delta Net] Add pallas training kernel to repo
#4376 opened Jul 8, 2026 by Rohan-Bierneni Collaborator Loading…
4 tasks done
1
Adding type suppressions for pyrefly
#4375 opened Jul 7, 2026 by copybara-service Bot Loading…
Move NNX state reconstruction and Linen conversion helpers to train_state_nnx.py.
#4374 opened Jul 7, 2026 by copybara-service Bot Loading…
1
Update Orbax dependency for MaxText CUDA 12.
#4373 opened Jul 7, 2026 by copybara-service Bot Loading…
1
[NNX] Emergency checkpoint: build the manager with a Linen-layout abstract
#4372 opened Jul 7, 2026 by ecnal-cienet Collaborator • Draft
4 tasks done
1
[WIP] Resize/pad VL decode works
#4370 opened Jul 7, 2026 by hengtaoguo Collaborator • Draft
4 tasks
1
Fix/nnx unroll scan axis
#4369 opened Jul 7, 2026 by hsuan-lun-chiang Collaborator Loading…
4 tasks done
7
Upgrade pre-training CI to jax 0.10.2
#4363 opened Jul 6, 2026 by gobbleturk Collaborator Loading…
4 tasks
1
Add DPO notebooks
#4362 opened Jul 6, 2026 by igorts-git Collaborator Loading…
4 tasks done
2
ProTip! What’s not been updated in a month: updated:<2026-06-08.