strisunshinewentingwang/pytorch

Pull Requests Commits

[LTC] Remove Constant op and class Literal (#68569)

alanwaketan committed 2 years ago

Verified 40f11f9e

[WIP] forbid aliasing between the outputs of a differentiable graph (#67732)

Krovatkin committed 2 years ago

ab1d879b

Revert D32283178: Add linalg.solve_triangular

janeyx99 committed 2 years ago

9f4e004a

Switch TsNode to the master version

desertfire committed 2 years ago

7ae2a999

[BC-breaking] Change dtype of softmax to support TorchScript and MyPy (#68336)

pearu committed 2 years ago

48771d1c

Switch BackendImplInterface and LoweringContext to the master version

desertfire committed 2 years ago

7dbbbd55

Revert D32187063: [static runtime] dequantize out variant

albanD committed 2 years ago

748d9d24

Add linalg.solve_triangular (#63568)

lezcano committed 2 years ago

0706607a

[static runtime] dequantize out variant (#67873)

ajyu committed 2 years ago

f1203356

Rename splitter result (#68303)

Shirong Wu committed 2 years ago

7d38768d

Fix DLPack CUDA stream convention (#67618)

Emilio Castillo committed 2 years ago

533e72e0

[testing] make @dtypes mandatory when using @dtypesIf (#68186)

kshitij12345 committed 2 years ago

d5d2096d

torch.linalg.qr: forward AD support (#67268)

nikitaved committed 2 years ago

857fed1f

[BE] MapAllocator: report map error on Linux (#68545)

malfet committed 2 years ago

a2d187a6

Fix `_make_wrapper_subclass`'s storage_offset handling (#68268)

zou3519 committed 2 years ago

b1aa45a8

Stop warning spamming about vmap in gradcheck (#68586)

zou3519 committed 2 years ago

f0e2ad50

Replace empty with new_empty in nn.functional.pad (#68565)

zou3519 committed 2 years ago

f9ef807f

[quant][embedding qat] eager mode QAT for Embeddings (#66429)

b-koopman committed 2 years ago

6c9cf5e6

[GPU host alloc] Fast path for size 0 malloc (#68532)

jaewonlee-fb committed 2 years ago

dbbb0247

[static runtime][dper] multi_env tests for static runtime: selective enable (#67467)

ajyu committed 2 years ago

4635f571

[reland] simplify init_from_local_shards API (#68021)

wanchaol committed 2 years ago

35712a8e

Sparse CSR: add `convert_indices_from_csr_to_coo` (#66774)

rok committed 2 years ago

952ca25d

Fix c10d TCP store with mutex (#68499)

jiayisuse committed 2 years ago

96ba2099

Enable desync root cause analysis for NCCL (#68310)

jiayisuse committed 2 years ago

146a7f68

`scatter_reduce` (#68115)

rusty1s committed 2 years ago

98077871

WIP iterate on the backend API for device mgmt

wconstab committed 2 years ago

608cb890

[fx2trt] add converter for acc_ops.hardtanh (#68550)

842974287 committed 2 years ago

e72b9db4

[predictor][trt] Expose more CUDA/CuDNN info to at::Context and BC stage 1 (#68146)

kflu committed 2 years ago

9d9ca88f

[android][fbjni] Update fbjni to 0.2.2 (#68400)

IvanKobzarev committed 2 years ago

d71092f6

[bugfix] TensorList args in functionalization pass (#68395)

bdhirsh committed 2 years ago

53bfb00e

Newer Older