pytorch/xla

Pull Requests Commits

Increase the cpp test timeout to 1000s

alanwaketan committed 2 years ago

55a08239

Increase the cpp test timeout to 1000s

alanwaketan committed 2 years ago

7e65d3d0

Update the commit

alanwaketan committed 2 years ago

d3ec960f

Revert local_repository

alanwaketan committed 2 years ago

c5bf94e1

pick a better commit

alanwaketan committed 2 years ago

3a329b8b

alanwaketan committed 2 years ago

f169beb1

alanwaketan committed 2 years ago

c0e88d87

[OpenXLA] Add openxla.md (#5507)

ManfeiBai committed 2 years ago

Verified 02c4aff4

Add PyTorch/XLA SPMD README.md (#5490)

yeounoh committed 2 years ago

Verified 4582675b

Enable megacore_dense by default (#5520)

alanwaketan committed 2 years ago

Verified 17a0b7c0

Modify kokoro to make it work (#5457)

qihqi committed 2 years ago

Verified 17513f74

Add dtype check to amp for tpu (#5505)

cowanmeg committed 2 years ago

Verified a06b1dff

[SPMD] Propagate replicated output (#5508)

alanwaketan committed 2 years ago

Verified 5c4ce2a9

mark_sharding over a replicated tensor is allowed. (#5513)

yeounoh committed 2 years ago

Verified 6e90569d

Get original_traced_args as example_inputs. (#5511)

qihqi committed 2 years ago

Verified 33f1cd29

Update dynamo.md (#5509)

shauheen committed 2 years ago

Verified 808015d8

Update dynamo doc (#5506)

JackCaoG committed 2 years ago

Verified 2f4ef46d

Fix tupled partition spec test on v3 (#5503)

jonb377 committed 2 years ago

Verified 81063fb8

remove torchvision in stablehlo.py (#5501)

lsy323 committed 2 years ago

Verified 1c42ba74

Add explict error message when tensor is on CPU for dynamo backend (#5499)

JackCaoG committed 2 years ago

Verified 9557fb72

Add a API to explictly init runtime (#5500)

JackCaoG committed 2 years ago

Verified e1ee592d

Support tuples in partition spec (#5488)

jonb377 committed 2 years ago

Verified 7a63c276

Refactor stablehlo API and put them in official location. (#5493)

qihqi committed 2 years ago

Verified 17abce76

Skip a couple tests on TPU due to precision issue (#5496)

JackCaoG committed 2 years ago

Verified d3d6ca30

Always use ExecuteReplicated with SPMD (#5494)

jonb377 committed 2 years ago

Verified 03991d44

Support input sharding changed after first dynamo tracing (#5477)

JackCaoG committed 2 years ago

Verified d1212e90

[SPMD] Make IR sharding custom sharding op (#5433)

alanwaketan committed 2 years ago

Verified dea9b296

[original author: mrnikwaws] Neuron operator support (#5471)

aws-kingrj committed 2 years ago

Verified 3e370b50

[SPMD] Patch nn.Linear (#5491)

alanwaketan committed 2 years ago

Verified 4f378e69

Add topological sorting to dynamo partitions (#5472)

wonjoo-wj committed 2 years ago

Verified b58327e1

Older