pytorch/xla

Pull Requests Commits

Fixed linalg.eigh by adding lowering and adding custom atol/rtol (#8072)

anishfish2 committed 1 year ago

Verified 855ec40e

[developer] Add a script to automate building torch_xla (#7876)

tengyifei committed 1 year ago

Verified 4e691dd7

Avoids error removing linalg.eig from skiplist by modifying atol and rtol (#8068)

anishfish2 committed 1 year ago

Verified 72fda76e

[torch_xla2] Fix op_info test for scatter_reduce (#8058)

dvhg committed 1 year ago

Verified 5b6e8cf1

Update trigger to 2.5.0-rc3 for 2.5 release (#8055)

ManfeiBai committed 1 year ago

Verified 22cf1976

add `resize_`, `resize_as_`, `rot90`, `repeat_interleave`, `combinations`, `diag_embed` (#8036)

ManfeiBai committed 1 year ago

Verified 47d231f3

Update tests for Neuron (#8053)

jeffhataws committed 1 year ago

Verified e657c872

[torch_xla2] Basic DDP example in a notebook (#8042)

will-cromar committed 1 year ago

Verified ea8c47a3

[torch_xla2] Fix nn.functional.conv2d and conv3d (#8048)

dvhg committed 1 year ago

Verified 6af550aa

[torch_xla2] Fix scatter and scatter_add (#8045)

dvhg committed 1 year ago

Verified 2aea97a7

parallel_loader: Fix TPU memory leak when calling __iter__. (#8039)

dudulightricks committed 1 year ago

Verified d79a37c4

Update and short circuit the Neuron initialization (#8041)

rpsilva-aws committed 1 year ago

Verified d0ea5cc2

Op info test for linalg.norm, linalg.pinv, linalg.solve (#7503) (#8044)

yenkwang committed 1 year ago

Verified 441098b7

Fix minor grammar error (#8043)

simonteozw committed 1 year ago

Verified c0501f01

Fixed masked_scatter and masked_select (#8037)

anishfish2 committed 1 year ago

Verified 52673541

Implement torch diagflat, and movedim as well as it is also needed as a helper function (#7378) (#8038)

simonteozw committed 1 year ago

Verified b79a46b0

Fix quantile, nanquantile and gather (#8040)

matinehAkhlaghinia committed 1 year ago

Verified 6b33f8f6

Uncomment tests that are passing. (#8035)

cornmander committed 1 year ago

Verified 0415cc60

Fix unbind op. (#8033)

cornmander committed 1 year ago

Verified 4fe55f55

Implement empty_like in torch aten ops (#8029)

nupurbaghel committed 1 year ago

Verified 0da723fc

Cache flash attention tracing (#8026)

JackCaoG committed 1 year ago

Verified 8b936115

Remove contiguity assertion from `XLATensorImpl`. (#7998)

ysiraichi committed 1 year ago

Verified 03374cd0

Fix op test for `nn.functional.scaled_dot_product_attention"nn.functi… (#8025)

guyao committed 1 year ago

Verified 3b0097d2

add `masked.amin` .. `masked.mean` (#8014)

ManfeiBai committed 1 year ago

Verified 0e477391

Update nightly version in README.md (#8024)

bhavya01 committed 1 year ago

Verified 5ec51397

[torchxla2] Op Info test for fft (#8018)

qihqi committed 1 year ago

Verified 3f4b31bb

Add fixes for narrow_copy and narrow ops (#8021)

matinehAkhlaghinia committed 1 year ago

Verified 008004d9

Implement dist in torch aten ops (#8019)

nupurbaghel committed 1 year ago

Verified 1b1eea21

Implement operations `cummax`, `cummin` (#8013)

guyao committed 1 year ago

Verified 2a5897d4

Update Openxla-pin to Sep13, libtpu-pin to Sep13, jax to 0.4.33_nightly (#7916)

ManfeiBai committed 1 year ago

Verified 37894b08

Newer Older