pytorch/xla

Pull Requests Commits

Delete .torch_pin (#2520)

ailzhang committed 5 years ago

Verified b68291a2

Add a job to make sure torch_pin is removed before landing. (#2521)

ailzhang committed 5 years ago

Verified 8c64d341

Remove celu registration since it's available in PyTorch core. (#2519)

ailzhang committed 5 years ago

Verified db18ba57

Use API file in new location, allows overriding composite ops. (#2510)

ailzhang committed 5 years ago

Verified b918dc4a

Add beta to the smooth_l1_loss lowering (#2496)

JackCaoG committed 5 years ago

Verified 5f5ecf18

Add a few out fns. (#2512)

ailzhang committed 5 years ago

Verified 1155541f

Fix VLOGs being disabled (#2509)

jysohn23 committed 5 years ago

Verified 7dfa77fd

Revert "Revert "Remove override for native_group_norm since we move the composite ker…"" (#2505)

ailzhang committed 5 years ago

Verified f32e1370

Add TriangularSolve to XLABuilder OP (#2507)

JackCaoG committed 5 years ago

Verified 132481bb

Bring in new tf (#2506)

JackCaoG committed 5 years ago

Verified b9a730fd

Remove true_divide in codegen as its lowering is removed. (#2503)

ailzhang committed 5 years ago

Verified 7139bc21

Revert "Remove override for native_group_norm since we move the composite ker…" (#2504)

ailzhang committed 5 years ago

Verified 716f44a9

Remove override for native_group_norm since we move the composite ker… (#2489)

ailzhang committed 5 years ago

Verified 80252658

Handle the int type for atan2 (#2500)

JackCaoG committed 5 years ago

Verified 157080a7

Asinh, Acosh, Atanh lowering (#2501)

JackCaoG committed 5 years ago

Verified f5184b82

Implement the var lowering (#2495)

JackCaoG committed 5 years ago

Verified 14d617a2

Updates div to perform true division in pt/xla (#2453)

JackCaoG committed 5 years ago

Verified 50b4f195

Cleanup debug images (#2486)

zcain117 committed 5 years ago

Verified 8d960a52

Dsiable bfloat16 i0 tests because of precision issue (#2485)

JackCaoG committed 5 years ago

Verified 6f486a24

Handle unordered reduce dimensions (#2484)

JackCaoG committed 5 years ago

Verified f03a7393

Simple setup docs and package name override to avoid conflicts when necessary (#2481)

cjolivier01 committed 5 years ago

Verified 8af57fb5

Only overwrite test precision if the pt/xla tolerance is larger (#2478)

JackCaoG committed 5 years ago

Verified 8b9f5c29

Update doc to build docker inside xla directory (#2474)

JackCaoG committed 5 years ago

Verified a2a53222

Remove float16 test from skip test since they should get auto skip (#2475)

JackCaoG committed 5 years ago

Verified 006cf17f

Remove registering fallthrough (moving to pytorch core). (#2463)

ailzhang committed 5 years ago

Verified c6082574

Cache debug docker and new conda install (#2476)

zcain117 committed 5 years ago

Verified d214e464

Add all_to_all back to our doc (#2472)

JackCaoG committed 5 years ago

Verified df0a2776

Implement the lowering of baddbmm op (#2471)

JackCaoG committed 5 years ago

Verified d2330676

Allow forwarding args to data loader and train/test function + allow "-" and "." in node name (#2470)

Chris Olivier committed 5 years ago

Verified ec79c5f4

Update tensorflow to include the all_to_all fix (#2469)

JackCaoG committed 5 years ago

Verified c712c93e

Newer Older