strisunshinewentingwang/pytorch

Pull Requests Commits

Adding support for comparing two bool varibales (#51844)

nikithamalgi committed 3 years ago

9c0caf03

[te] Benchmark vml-based logit (#51771)

bertmaher committed 3 years ago

602434bc

[te] Implement log approximation using the VML approach (#51752)

bertmaher committed 3 years ago

2e35fe95

[te] Introduce likely/unlikely CompareSelect hint (#51751)

bertmaher committed 3 years ago

ff73be7e

[te][llvm] Generate arithmetic vs logical right shift as appropriate (#51749)

bertmaher committed 3 years ago

74082f0d

Back out "Revert D26009829: Optimize relu on cpu using clamp_min" (#51819)

bertmaher committed 3 years ago

0620c96f

fake_quant cachemask: remove Python bindings (#51878)

vkuzo committed 3 years ago

33afb5f1

[model loading] Add max_batch_size override for batch size exploration

ChunliF committed 3 years ago

5f9fb93c

Migrate masked_fill__cuda to ATen (#51404)

kshitij12345 committed 3 years ago

76866291

ns_eager: rename Logger I/O var names to logger_cls (#51359)

vkuzo committed 3 years ago

929b91a2

Automated submodule update: FBGEMM (#52014)

facebook-github-bot committed 3 years ago

5a9bac58

add more logging fields that can be set in construction time (#51260)

zhaojuanmao committed 3 years ago

18e0a613

[FX] Generalize dict key check in `create-arg` (#51927)

James Reed committed 3 years ago

d23cb940

[FX][EZ] Fix tuple type annotations (#52010)

James Reed committed 3 years ago

256f93fb

[FX] Fix leaf modules in Transformer (#51998)

James Reed committed 3 years ago

d4e84b0c

[PyTorch] Re-order TensorImpl fields to save a word (#50920)

swolchok committed 3 years ago

d5a9627f

[FX] Make some modifications to limitation section (#51928)

Chillee committed 3 years ago

475278f1

Let child CUDAFuture wait for parent CUDAFuture's CUDAEvents (#51820)

mrshenli committed 3 years ago

3af7b673

[ROCm] add 4.0.1 docker image (#51507)

jeffdaily committed 3 years ago

c6b4fc8a

[DataLoader] Rename files of functional datapipes (#51880)

ejguan committed 3 years ago

1921b244

[DataLoader] Rename Callable to Map IterDataPipe (#51879)

ejguan committed 3 years ago

9eb70c3c

[DataLoader] Implement FilterIterDataPipe (#51783)

ejguan committed 3 years ago

104371e1

[pytorch] recast infer_type error and amend with name and item that failed inferring

Aapo Kyrola committed 3 years ago

e964d77f

Fixing Softmax bench. (#51898)

navahgar committed 3 years ago

12d85b53

[C2] Add shape inference logic for ColwiseMax operator. (#51914)

kennyhorror committed 3 years ago

7e54a648

[FX] make map_arg require a callable (#51907)

jansel committed 3 years ago

0410cba2

[Pytorch Mobile] Only preserve bundled input helpers for forward if they exist (#51884)

JacobSzwejbka committed 3 years ago

2f2b1700

Fix the lifetime of PyTensorType (#51649)

antocuni committed 3 years ago

8fab33f9

[docs] Add docs for storage and tensors for quantized Tensor (#51817)

jerryzh168 committed 3 years ago

0ec00c12

Make RebatchingBuffer compatible with auto shape inference

Kai Yang committed 3 years ago

fc314350

Newer Older