strisunshinewentingwang/pytorch

Pull Requests Commits

Publish all_gather_object and gather_object docs (#43772)

mrshenli committed 4 years ago

2f527485

Revert D23385091: [quant][graphmode][fx] Add top level APIs

albanD committed 4 years ago

f7bae5b6

Revert D23385090: [quant][graphmode][fx] Add support for weight prepack folding

albanD committed 4 years ago

68304c52

[fix] torch.multinomial : fix for 0 size dim (#43775)

kshitij12345 committed 4 years ago

0394c5a2

Update aliasing in tensorexpr fuser (#43743)

Elias Ellison committed 4 years ago

3c8b1d73

use types in the IR instead of vmap (#43742)

Elias Ellison committed 4 years ago

5da8a7bf

Add passes to profiling executor pipeline (#43636)

Elias Ellison committed 4 years ago

259e5b7d

Use prim::TensorExprGroup interned symbol (#43635)

Elias Ellison committed 4 years ago

a7e7981c

Update requires grad property (#43634)

Elias Ellison committed 4 years ago

1c0faa75

add qr_backward functionality for wide case (#42216)

rlucas7 committed 4 years ago

2bede78a

[RPC profiling] Add test to ensure using record_function works for RPC (#43657)

rohan-varma committed 4 years ago

69dd0bab

Add __complex__ (#43844)

zasdfgbnm committed 4 years ago

4ef12be9

addmm/addmv should accept complex alpha and beta (#43827)

zasdfgbnm committed 4 years ago

c5d0f091

[fx] GraphModule.src -> GraphModule.code (#43655)

suo committed 4 years ago

89452a67

[NNC] Hook up registerizer to Cuda codegen [2/x] (#42878)

nickgg committed 4 years ago

1390cad2

Better msg (#43848)

Yinghai Lu committed 4 years ago

63dbef30

[quant][graphmode][fx] Add support for weight prepack folding (#43728)

jerryzh168 committed 4 years ago

ef08f920

[quant][graphmode][fx] Add top level APIs (#43581)

jerryzh168 committed 4 years ago

eb4199b0

Properly check that reduction strings are valid for l1_loss, smoothl1_loss, and mse_loss. (#43527)

gchanan committed 4 years ago

42c895de

[quant][graphmode][fx][fix] enable per channel quantization for functional ops (#43534)

jerryzh168 committed 4 years ago

b8d34547

Rewrite of ATen code generator (#42629)

ezyang committed 4 years ago

6ea89166

Print all traceback for nested backwards in detect_anomaly (#43626)

mfkasim1 committed 4 years ago

576880fe

Test runner for batched gradient computation with vmap (#43664)

zou3519 committed 4 years ago

1cdb9d2a

Kill unused _pointwise_loss function. (#43523)

gchanan committed 4 years ago

1dcc4fb6

[resubmit] Add amax/amin (#43819)

zasdfgbnm committed 4 years ago

a860be89

Enable complex blas for ROCm. (#43744)

jeffdaily committed 4 years ago

8fb7c502

[ONNX] Utilize ONNX shape inference for ONNX exporter (#40628)

BowenBao committed 4 years ago

08126c91

Documents sub properly, adds subtract alias (#43850)

Mike Ruberry committed 4 years ago

3aeb70db

Disable RocM CircleCI jobs (#42630)

malfet committed 4 years ago

3dc96454

Update CUDA11 docker container (#42200)

zasdfgbnm committed 4 years ago

7b835eb8

Newer Older