strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "Add torch.float8_e5m2 and torch.float8_e4m3 data types (#104242)"

pytorchmergebot committed 1 year ago

f2b15772

[C10D] Improve MTPG autograd test. Fixes #105106 (#105356)

Rodrigo Kumpera committed 1 year ago

02cd971e

Improved error messages for deprecated linalg functions. (#105506)

franklinscudder committed 1 year ago

ded9b942

Enable intellisense for _dynamo, _inductor and onnx by importing under type_checking guard (#105361)

BowenBao committed 1 year ago

ca126880

Add torch.float8_e5m2 and torch.float8_e4m3 data types (#104242)

australopitek committed 1 year ago

a9804130

Unify TEST_CUDNN definition (#105594)

ezyang committed 1 year ago

1ea153a1

Rely on is_expr_static_and_true to test gcd (#105578)

ezyang committed 1 year ago

692e0566

[inductor] Fix an AOTInductor missing output issue (#105496)

desertfire committed 1 year ago

71067631

inductor: add support for 0 repeats (#105446)

blzheng committed 1 year ago

0e4c1215

[inductor] fix duplicate arg handling in triton templates (#105315)

constroy committed 1 year ago

0578732b

Remove unnecessary left == right test. (#105576)

ezyang committed 1 year ago

a5317ae8

[ONNX] Suppress ORT warnings in unit tests (#105624)

BowenBao committed 1 year ago

980589b0

[Inductor] aten.dist decomposition (#105586)

yanboliang committed 1 year ago

8daed86e

Revert "inductor: promote half/bfloat16 constant to float for cpu vectorization path (#105440)"

pytorchmergebot committed 1 year ago

dfc98747

[quant][pt2e][be] Rename qnnpack quantizer to xnnpack quantizer (#105551)

jerryzh168 committed 1 year ago

dff4e034

Back out "Make adding buffers more like adding parameters (#104069)" (#105581)

atalman committed 1 year ago

c6653b65

Remove dead sizevars.__getitem__ (#105579)

ezyang committed 1 year ago

2e81cdc1

Tighten size_hint invariant (#105580)

ezyang committed 1 year ago

43540a1c

Enable more e2e foreach optimizer compilation tests (#105438)

mlazos committed 1 year ago

690ea933

Reland: Value range refinement using multi-variate expressions (#105491)

ysiraichi committed 1 year ago

0cd51b3d

[PyTorch] [Memory profiler] Early return if qualified name is invalid (#105495)

chowarfb committed 1 year ago

3dacc8e8

Update slow CI jobs to rocm5.6 (#105516)

jithunnair-amd committed 1 year ago

86076abe

[PyTorch] Remove device transfers from JNI (#105583)

jorgep31415 committed 1 year ago

93e6fc54

Reenable UFMT on pyi (#105577)

ezyang committed 1 year ago

0b524343

[inductor] Disable cudagraphs if index_put_ fallback is encountered (#105439)

davidberard98 committed 1 year ago

5abc5ab5

[XNNPACK][QS8] torch.slice (#105252)

mcr229 committed 1 year ago

bc6bca9d

[Quant][PT2E] Remove x86 inductor pt2e backend config (#105039)

leslie-fang-intel committed 1 year ago

fa6be2fa

[dynamo][rewrite_asserts] Insert assertion msg in bytecode only when needed (#105549)

anijain2305 committed 1 year ago

af9a4e08

[Quant][Inductor] Use truncate instead of default rounding round when convert float to uint8 (#105109)

leslie-fang-intel committed 1 year ago

6c432381

Migrate tuple(handle) -> handle (#104488)

voznesenskym committed 1 year ago

a8329676

Newer Older