strisunshinewentingwang/pytorch

Pull Requests Commits

[PyTorch][CUDA Caching Allocator] Export sync-stream-and-free-HBM counter in memory_stats for performance debugging (#120050)

levythu committed 207 days ago

b6139b1e

[executorch hash update] update the pinned executorch hash (#120675)

pytorchupdatebot committed 207 days ago

a1c641f1

Restore artifact name in log messages (#120671)

ezyang committed 207 days ago

23777313

[vision hash update] update the pinned vision hash (#119944)

pytorchupdatebot committed 207 days ago

ac285717

[audio hash update] update the pinned audio hash (#120135)

pytorchupdatebot committed 207 days ago

9d423f0e

[dynamo][guards-cpp-refactor] DictGetItemGuardAccessor for f_locals (#120593)

anijain2305 committed 207 days ago

63f874b4

docker: Only match tags that start with v* (#120670)

seemethere committed 207 days ago

27990045

Remove hard numpy dependency from experimental_ops.py (#119520)

cpuhrsch committed 207 days ago

cf6df886

Switch to native functional collective by default (#120370)

yifuwang committed 207 days ago

2de7468d

[dynamo][compile-time] Collect guard debug stack info only with logs enabled (#120520)

anijain2305 committed 207 days ago

8a59f49d

[EZ][BE] Use nested namespace in functorch (#120663)

malfet committed 207 days ago

2e0e5457

[1/2] Intel GPU Runtime Upstreaming for Generator (#118528)

guangyey committed 207 days ago

b3fe53e1

Add torch.ops.aten.print (#120295)

angelayi committed 207 days ago

f064dec7

Replace individual detaches with overall torch.no_grad decorator (#120638)

janeyx99 committed 207 days ago

ef9b6d68

clip_grad_norm can use fast foreach path for inf norm (#120623)

janeyx99 committed 207 days ago

df72819f

Revert "Add torch.ops.aten.print (#120295)"

pytorchmergebot committed 207 days ago

b01bd1f7

Revert "[Dynamo] Remove deadcode: unwrapping script_if_tracing (#120444)"

pytorchmergebot committed 207 days ago

17560eb4

Mark test_reference_numerics_extremal__refs_frexp_cuda as xfail on windows (#120640)

isuruf committed 207 days ago

e874376f

Revert [dynamo] support group=None when rewriting collectives (#12018) (#120677)

kit1980 committed 207 days ago

d341b66e

[meta registration] efficient_attention_forward fix for NT inputs (#120594)

davidberard98 committed 207 days ago

fdae9363

Add structured trace logs (#120289)

ezyang committed 207 days ago

9dfaef96

[dynamo] fix segfault in _debug_get_cache_entry_list (#120635)

williamwen42 committed 207 days ago

ecb3f33a

Add the hyperlink of the transfomer doc (#120565)

lancerts committed 207 days ago

64660b51

Implement aten::upsample_linear1d on mps (#115031)

kaieberl committed 207 days ago

c59b1416

Add cpp stack traces to our own reruns (#119408)

albanD committed 207 days ago

30625ae5

Revert "Switch to native functional collective by default (#120370)"

pytorchmergebot committed 207 days ago

41adec3c

Use lxml in scripts/compile_tests when it is available (#120633)

zou3519 committed 207 days ago

7b1cc140

[Dynamo] Fix guards for script_if_tracing or lru_cache fn with default args (#120390)

yanboliang committed 207 days ago

5a0a9644

[PT2][Inductor]Add unbind node normalization (#120253)

mengluy0125 committed 207 days ago

55b59084

[FSDP] Removed `.detach` in `clip_grad_norm_` (#120612)

awgu committed 207 days ago

274b3624

Newer Older