Skip to content

Pull requests: jax-ml/jax

New pull request New
567 Open 21,991 Closed
567 Open 21,991 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

Integrate Triton up to
#31739 opened Sep 11, 2025 by copybara-service bot Loading…
[Mosaic:TPU][Relayouts] Replicated minor large-to-small retiling (one-to-many)
#31730 opened Sep 11, 2025 by copybara-service bot Loading…
Fix ann_test.py numerical bug in target reshape
#31720 opened Sep 10, 2025 by gulsumgudukbay Loading…
8
Remove pltpu.MemorySpace.ANY
#31688 opened Sep 10, 2025 by copybara-service bot Loading…
@sharadmv
[pallas:mosaic_gpu] Sketched a decorator exposing Pallas Mosaic GPU kernels to PyTorch
#31663 opened Sep 9, 2025 by copybara-service bot Loading…
@superbobry
[Mosaic:TPU][NFC] Simplify computeSqueezeDimsChecked
#31652 opened Sep 9, 2025 by copybara-service bot Loading…
Enhance TPU Matmul RHS transpose fusion to handle intervening slices.
#31625 opened Sep 8, 2025 by copybara-service bot Loading…
Move device_put and shard_args into their own file.
#31576 opened Sep 4, 2025 by copybara-service bot Loading…
@hawkinsp
[CI] Testing integration with skipped test
#31570 opened Sep 4, 2025 by copybara-service bot Loading…
@MichaelHudgins
Fix Jax OSS CI CI Optional GPU Presubmit Label to flag PR to run additional GPU testing not in standard presubmits
#31562 opened Sep 4, 2025 by copybara-service bot Loading…
Integrate Triton up to [7decd434](https://github.com/openai/triton/commits/7decd434a2dbf76e74ea3c2a5ec877326b83f4fa)
#31554 opened Sep 4, 2025 by copybara-service bot Loading…
[pallas:mgpu] emit_pipeline_emit_pipeline to support different number of stages for each input.
#31543 opened Sep 4, 2025 by copybara-service bot Loading…
Remove size scaling of tolerances in test_util.
#31528 opened Sep 3, 2025 by copybara-service bot Loading…
@jakevdp
Make roofline BUILD target & add call_tf.
#31524 opened Sep 3, 2025 by copybara-service bot Loading…
Reverts cddb59e9a93882395284281e42119a411360608d
#31523 opened Sep 3, 2025 by copybara-service bot Loading…
Reverts 71fe0c3e3fbf5211303dce6f3e341f13f5c3551a
#31522 opened Sep 3, 2025 by copybara-service bot Loading…
Reverts 71fe0c3e3fbf5211303dce6f3e341f13f5c3551a
#31519 opened Sep 3, 2025 by copybara-service bot Loading…
[pallas:mgpu] emit_pipeline to support different number of stages for each input.
#31513 opened Sep 3, 2025 by copybara-service bot Loading…
Reverts changelist 796974894
#31489 opened Sep 2, 2025 by copybara-service bot Loading…
[pallas:mgpu] warp specialized pipeline with special delay_release.
#31483 opened Sep 2, 2025 by copybara-service bot Loading…
Adds DeserializeExecutable Overloads for the new FfiHostCallback code path.
#31469 opened Sep 1, 2025 by yliu120 Loading…
@emilyfertig
1
[Pallas:MGPU] Add a primitive to reduce TCGEN05 to WGMMA for interleaved 0:16 data input
#31460 opened Sep 1, 2025 by copybara-service bot Loading…
Canonicalize to vclamps (for vllm fp8 quant, maybe other uses)
#31452 opened Aug 31, 2025 by copybara-service bot Loading…
Handle overlapping specs in scaled matmul custom partitioning
#31433 opened Aug 29, 2025 by jaro-sevcik Loading…
[Mosaic:TPU] Better tpu.memref_reshape support for packed types by improving the reshaped layout inference
#31418 opened Aug 28, 2025 by copybara-service bot Loading…
ProTip! Type g i on any issue or pull request to go back to the issue listing page.