strisunshinewentingwang/pytorch

Pull Requests Commits

Correct LBFGS tolerance_grad doc string (#99792)

arthurfeeney committed 1 year ago

7a8d0ccd

Preserve mark_dynamic when cloning inputs (#99617)

ezyang committed 1 year ago

f602b3a6

Add memoized_only mode to tensor conversion (#99741)

voznesenskym committed 1 year ago

5e73569a

Guard static shapes alongside tensors, instead of from shape_env, in dynamic_shapes=True (#99566)

voznesenskym committed 1 year ago

4c289294

Fix torch.compile() on a skipped module (#98894)

jansel committed 1 year ago

220712f4

inductor: fix AllenaiLongformerBase dynamic shape error on CPU (#98842)

blzheng committed 1 year ago

d192729c

[dynamo] disallow_in_graph bugfix (#99600)

anijain2305 committed 1 year ago

31eb9949

[Executorch][XNNPACK] Quantized Max Pool 2d (#99587)

mcr229 committed 1 year ago

e63c502b

Remove deprecated declaration suppression (#99749)

malfet committed 1 year ago

7749eec8

[quant][pt2e] add all convs-relu fusion qat configs (#99586)

mcr229 committed 1 year ago

a964a3db

[quant][pt2e] add dropout to executorch backend config (#99585)

mcr229 committed 1 year ago

c139dfd7

[spmd] Add list handling to data parallel and add foreach tests (#99373)

wanchaol committed 1 year ago

9db69206

[dtensor] add StrategyType and TupleStrategy (#99435)

wanchaol committed 1 year ago

c1e2fa81

[fx] Add a function to allow adding more functions to the side effect function set (#97288)

842974287 committed 1 year ago

82a54513

[Profiler] Support HTML plot output for profiler export_memory_timeline API (#99751)

aaronenyeshi committed 1 year ago

87b71e57

[BE][1/N]Add sharding spec logger for ShardedTensor (#99748)

wz337 committed 1 year ago

ca8625f4

[ONNX] Add additional_test_kwargs into test_fx_to_onnx_with_onnxruntime.py (#99434)

titaiwangms committed 1 year ago

bd719111

[spmd] Introduce Compile Mode FSDP with DTensor (#99062)

wanchaol committed 1 year ago

e9bf9414

[vision hash update] update the pinned vision hash (#99486)

pytorchmergebot committed 1 year ago

be62a807

[ONNX] Support aten::scaled_dot_product_attention in torchscript exporter (#99658)

titaiwangms committed 1 year ago

e5664c65

[docs] nn.functional.embedding: Note expected discrepancy between numerical and analytical gradients (#99181)

soulitzer committed 1 year ago

6585d76f

lower cmake version requirement in FindSanitizer.cmake (#97073)

cyyever committed 1 year ago

2b7161e2

Add pattern to normalize split (#99588)

devashishshankar committed 1 year ago

93d0a9c1

Add pass to remove redundant conversions (#99697)

jansel committed 1 year ago

79ec91a9

Revert "Simplify _use_grad_for_differentiable (#98706)"

pytorchmergebot committed 1 year ago

4637c5ae

[ONNX] Cover 'undiscoverable' ops 'torch.ops.aten' (#99682)

BowenBao committed 1 year ago

872319d3

Discover and run C++ tests with run_test.py (#99559)

huydhn committed 1 year ago

96d3f3de

adjust batch sizes for hf suite (#99691)

Natalia Gimelshein committed 1 year ago

bfbc4e74

[BE][DTensor] validate the mesh argument in DeviceMesh construction (#99094)

XilunWu committed 1 year ago

ce609973

Allow torch.fx to take Modules that return dataclass (#99576)

idning committed 1 year ago

cf357adc

Newer Older