pytorch/xla

Pull Requests Commits

JackCaoG committed 1 year ago

e678cc8d

Add docs for eager mode

JackCaoG committed 1 year ago

2eec5039

Update tuple type annotation (#7699)

JackCaoG committed 1 year ago

Verified e523fdf5

Add atan2, arange, addcmul, addcdiv, _softmax_backward_data, T, H (#7681)

ManfeiBai committed 1 year ago

Verified a69fd510

Optimize execution for ops that have multiple output in eager mode (#7680)

JackCaoG committed 1 year ago

Verified b2c7f65e

fix git fetch for torch pin checkout (#7690)

JackCaoG committed 1 year ago

Verified 9cf1fd8e

Lower aten::_linalg_eigh (#7674)

tengyifei committed 1 year ago

Verified f975ad64

Extend timeout for `tpu-info` test (#7686)

will-cromar committed 1 year ago

Verified 3bbd3f65

Update 2.4 rc-8 in build trigger (#7685)

bhavya01 committed 1 year ago

Verified 8fdcb952

Delete extra file (#7683)

will-cromar committed 1 year ago

Verified 08e63e32

Introduce torch_xla.launch() (#7648)

zpcore committed 1 year ago

Verified 5b8e8e01

Fix helm version format in terraform (#7678)

will-cromar committed 1 year ago

Verified 75b10b7f

Update TPU ARC cluster (#7677)

will-cromar committed 1 year ago

Verified 8390de23

support gmm as a custom op for dynamo (#7672)

JackCaoG committed 1 year ago

Verified 1651e764

Update openxla pin to 11 July 2024 (#7639)

bhavya01 committed 1 year ago

Verified 8043050f

Replace remaining usages of `jax.lax.Precision` (#7671)

will-cromar committed 1 year ago

Verified ac26a977

Add shard_map wrapper (#7670)

lsy323 committed 1 year ago

Verified 18ed9dd8

[Fori_loop] Update `randint` max range to Support `bool` dtype (#7632)

huzama committed 1 year ago

Verified 41d998de

Use jax.export insteadl of jax.experimental.export (#7300)

qihqi committed 1 year ago

Verified 34736f0f

lower _aten_constant_pad_nd (#7613)

wenxindongwork committed 1 year ago

Verified b32a91f2

Add `vsplit`, `vstack`, `vdot` ops (#7664)

miladm committed 1 year ago

Verified 66ff604a

[torch_xla2] Functional collectives (#7580)

will-cromar committed 1 year ago

Verified 3078d163

Asymmetric quantized matmul support (#7626)

lsy323 committed 1 year ago

Verified 289471cb

Easier way to make img show up in our doc page (#7644)

JackCaoG committed 1 year ago

Verified 72069750

Move tensor constructors to `jaten` (#7615)

will-cromar committed 1 year ago

Verified 5e9da938

[torch_xla2] Add broadcast_to, isin, item, lcm, lerp (#7638)

ManfeiBai committed 1 year ago

Verified ccd3da2c

[torch_xla2] Add fmax, fmin, frexp (#7603)

ManfeiBai committed 1 year ago

Verified c62b1d46

Add blockwise quantized dot support (#7605)

lsy323 committed 1 year ago

Verified 88bcb45f

Update `apt-get update` in upstream image build (#7640)

will-cromar committed 1 year ago

Verified edd24532

separate PyTorch/XLA docs (#7593)

JackCaoG committed 1 year ago

Verified 275bf919

Older