pytorch/xla

Pull Requests Commits

Add block sizes for Qwen/Qwen2.5-32B-Instruct (#9516)

vanbasten23 committed 96 days ago

Verified 531c724f

Error Handling: return status value when loading PjRt dynamic plugin. (#9495)

ysiraichi committed 96 days ago

Verified 2820f7c5

Optimize w8a8 kernel vmem limit (#9508)

kyuyeunk committed 99 days ago

Verified 29ae4c76

Move mutable properties of env to thread local, misc changes (#9501)

qihqi committed 100 days ago

Verified 0a1594a8

Clean up quantized matmul condition code (#9506)

kyuyeunk committed 100 days ago

Verified 16b12022

Implement XLAShardedTensor._spec and test (#9488)

aws-cph committed 100 days ago

Verified ca471983

Error Handling: refactor `ComputationClient::TransferFromDevice` to propagate status. (#9429)

ysiraichi committed 100 days ago

Verified 299a16b5

[Bugfix] fix ragged attention kernel auto-tuning table key (#9497)

yaochengji committed 101 days ago

Verified 31c4c2f2

Fix python 3.11 cuda wheel link in the readme (#9493)

vfdev-5 committed 103 days ago

Verified e82631ec

Misc fixes: (#9491)

qihqi committed 104 days ago

Verified 26def0f2

Fix nested stableHLO composite regions (#9385)

Carlomus committed 104 days ago

Verified 55b7d021

Allow mixed tensor type math if one of them is a scalar (#9453)

qihqi committed 106 days ago

Verified 95ba754a

Remove support for one-process-per-device style of distributed. (#9490)

qihqi committed 106 days ago

Verified f3c7907f

Update README.md to reflect supported python versions (#9484)

bhavya01 committed 106 days ago

Verified 86a99d75

Add support for callable in torchax.interop.JittableModule.functional_call in the first parameter (#9451)

zmelumian972 committed 106 days ago

Verified ca914452

Implement collective gather op (#9435)

bfolie committed 107 days ago

Verified a021bf09

implement collective all_to_all op (#9442)

bfolie committed 107 days ago

Verified e7dcc7bd

Rename `XLA_SHOW_CPP_ERROR_CONTEXT` to `TORCH_SHOW_CPP_STACKTRACES` (#9482)

ysiraichi committed 107 days ago

Verified 58592c48

Improve error message for shape promotion on lowering. (#9486)

ysiraichi committed 107 days ago

Verified 2c556e73

Add r2.9 release nightly tests to README (#9478)

pgmoka committed 107 days ago

Verified 8b95c5d6

Fix spmd sharding visualization when device index is >= 10 (#9475)

jeffhataws committed 108 days ago

Verified c6aa90f1

Update Setuptools version in build dependencies. (#9487)

bhavya01 committed 108 days ago

Verified b612613d

Fix duplicate labels and other docs build warnings (#9446)

melissawm committed 108 days ago

Verified d7c9e506

Bump version (#9480)

qihqi committed 108 days ago

Verified 4b3d34da

Error Handling: refactor the existing `ComputationClient` implementations to use status QOL functions. (#9420)

ysiraichi committed 108 days ago

Verified 866fd22d

Optimize w8a8 pallas kernel (#9473)

kyuyeunk committed 109 days ago

Verified 36ff641d

Torchax: Allow reuse of the jittable procedure with different functio… (#9374)

zmelumian972 committed 109 days ago

Verified ebf9a8be

Update TPU CI with latest docker container (#9483)

bhavya01 committed 109 days ago

Verified ca04583b

Add dtensor mesh conversion test (#9474)

aws-cph committed 109 days ago

Verified 93a5e583

implement collective reduce op (#9437)

bfolie committed 110 days ago

Verified 173a7708

Newer Older