pytorch/xla

Pull Requests Commits

Syncfree SGD optimizer for AMP (#3145)

ymwangg committed 4 years ago

Verified 5174f9ff

Replace Usages of xla:util::Hash* with torch::lazy::Hash (#3148)

wconstab committed 4 years ago

Verified 52be55d2

Revert "Revert "Update _(log_)softmax_backward parameters (#3137)" (#3146)" (#3150)

JackCaoG committed 4 years ago

Verified 457270c2

Update doc link (#3151)

JackCaoG committed 4 years ago

Verified e3e38164

Nll loss update (#3144)

JackCaoG committed 4 years ago

Verified 13597d69

Move PT_XLA_DEBUG to the front (#3149)

JackCaoG committed 4 years ago

Verified 01bd3dd1

Revert "Update _(log_)softmax_backward parameters (#3137)" (#3146)

JackCaoG committed 4 years ago

Verified 419417dc

Update _(log_)softmax_backward parameters (#3137)

soulitzer committed 4 years ago

Verified 462a48ae

.circleci: Don't overwrite CMAKE_PREFIX_PATH (#3106)

seemethere committed 4 years ago

Verified e0e5f937

[POC] Add AMP zero gradients optimization (#3119)

ymwangg committed 4 years ago

Verified 3e7c301b

change master to main (#3139)

miaoshasha committed 4 years ago

Verified 8312a420

Skip AdaptiveMaxPool2D cpp test until tile mismatch bug is fixed on TPU (#3135)

JackCaoG committed 4 years ago

Verified 96d5e18e

shauheen committed 4 years ago

11d7c405

fixed TensorType for TPU device execution, limited the unit test chec… (#3116)

miladm committed 4 years ago

Verified 088607f2

match an optional 3rd digit for version number (#3128)

miaoshasha committed 4 years ago

Verified 5a75ff88

Enable CPP dynamic shape test on CPU and GPU (#3112)

JackCaoG committed 4 years ago

Verified 81da6008

Fix env var checking in test_data_type (#3101)

JackCaoG committed 4 years ago

Verified 01e772a4

test backward() calls don't hit the CPU fallback (#3092)

bdhirsh committed 4 years ago

Verified 6c2fc734

Add lowering for at::nan_to_num (#3093)

ymwangg committed 4 years ago

Verified bbbcbcd3

fix a minor doc typo (#3096)

thsunkid committed 4 years ago

Verified 905f9512

Grammatically updated API_GUIDE.md file by adding a comma under xla tensor deep dive section (#3097)

himanshu-0611 committed 4 years ago

Verified 0bb61d47

Rollback tf to pass internal wheel build (#3094)

JackCaoG committed 4 years ago

Verified 61e3960c

updating BN forward inference running_stats output (#2961)

jjsjann123 committed 4 years ago

Verified cb168ed7

Sgn implementation (#3045)

miladm committed 4 years ago

Verified 8d6b731d

Lower Logical_not/xor/or (#3084)

JackCaoG committed 4 years ago

Verified bbd87ad5

Add more debugging message to test_data_type (#3091)

JackCaoG committed 4 years ago

Verified a52a1325

update op lowering guide, update scatter ops to lower out-of-place in… (#3087)

bdhirsh committed 4 years ago

Verified 28c9cc45

Adaptive max pool (#3083)

JackCaoG committed 4 years ago

Verified 7ad90d6c

Run test_autocast for cpu and gpu on circleCI (#3089)

JackCaoG committed 4 years ago

Verified 4df13bfc

Minor reorder of std_mean member (#3085)

JackCaoG committed 4 years ago

Verified 08b5a744

Newer Older