pytorch/xla

Pull Requests Commits

Update artifacts_builds.tf for 2.8.0-rc3 wheel

bhavya01 committed 92 days ago

Verified 16f7fe73

Update artifacts_builds.tf for 2.8.0-rc2 (#9522)

bhavya01 committed 93 days ago

Verified 241cd47f

[XLA] Add placements property to XLAShardedTensor for DTensor compatibility (#9509)

Hoomaaan committed 94 days ago

Verified 95bee8f0

Deduplicate `GetXlaTensors()` function. (#9518)

ysiraichi committed 94 days ago

Verified cb64f4c7

Add w8a8 kernel blocks for Qwen 2.5 7B (#9517)

kyuyeunk committed 94 days ago

Verified 199a9bd5

Dump C++ and Status propagation stacktraces. (#9492)

ysiraichi committed 94 days ago

Verified 7aa466e2

Error Handling: refactor `GetXlaTensor` and related functions to use status types. (#9510)

ysiraichi committed 95 days ago

Verified cd3bd91f

Error Handling: refactor `ExecuteComputation` and `ExecuteReplicated` to propagate status. (#9445)

ysiraichi committed 95 days ago

Verified b0ffc491

Error Handling: propagate status for `ReleaseGilAndTransferData` and `XlaDataToTensors`. (#9431)

ysiraichi committed 95 days ago

Verified 1ed6b46c

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

vanbasten23 committed 95 days ago

Verified 531c724f

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

ysiraichi committed 95 days ago

Verified 2820f7c5

Optimize w8a8 kernel vmem limit (#9508)

kyuyeunk committed 98 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 106 days ago

Verified a021bf09

implement collective all_to_all op (#9442)

bfolie committed 106 days ago

Verified e7dcc7bd

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

ysiraichi committed 106 days ago

Verified 58592c48

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

ysiraichi committed 106 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 107 days ago

Verified c6aa90f1

Older