strisunshinewentingwang/pytorch

Pull Requests Commits

Update nn.Module._apply to not gate on should_use_set_data when swap_tensors is set (#120659)

mikaylagawarecki committed 206 days ago

677e67c3

[BE] fail_* variables don't need to be shared across restarts, they're set only once (#120712)

ezyang committed 206 days ago

213b3ac3

Add test to check that COW inputs are not materialized (#119507)

kurtamohler committed 206 days ago

2ebf2c88

Avoid COW materialization in `at::parallel_for/parallel_reduce` (#120455)

kurtamohler committed 206 days ago

cabc09a5

[Clang-tidy header][20/N] Fix clang-tidy warnings in aten/src/ATEN/*.{cpp,h} (#120574)

cyyever committed 206 days ago

1e9fafc1

use condition_variable and wait_until in nccl dump on timeout (#120544)

jeffdaily committed 206 days ago

9c597ff1

Fix broken link in README (#120698)

keenranger committed 206 days ago

14b258b5

[CUDA][cuBLAS] Check if a context is present when grabbing a cuBLAS handle (#120131)

eqy committed 206 days ago

5929d4e8

Revert "[inductor][Gemm] Autotune with matrix_instr_nonkdim for AMDGPU (#120639)"

pytorchmergebot committed 206 days ago

f36e00b8

Limit loop unrolling (#120023)

aorenste committed 206 days ago

6cc7f9a2

Revert "Add structured trace logs (#120289)"

pytorchmergebot committed 206 days ago

f3dd2a54

[CUDA][cuBLAS] Bump `test_cublas_baddbmm_large_input` tolerances (#117889)

eqy committed 206 days ago

65efece3

[dynamo] Add some helpers to PyCodegen (#120684)

jansel committed 206 days ago

5b5c167a

[dtensor] standardize tuple strategy handling for foreach ops (#120695)

wanchaol committed 206 days ago

0c8bb6f7

[profiler] log process group config information in distributedInfo field (#119443)

shengbao-zheng committed 206 days ago

440a9b21

[inductor][Gemm] Autotune with matrix_instr_nonkdim for AMDGPU (#120639)

htyu committed 206 days ago

78f53a3f

[export] Use forward hooks to capture module signatures. (#120468)

zhxchen17 committed 206 days ago

3f62b05d

Add lowering for adaptive_max_pool2d (#120254)

amjames committed 206 days ago

ed3c256b

[AOTI] Fix a strict-aliasing warning (#120628)

desertfire committed 207 days ago

27bb73fe

[inductor] correctly retrieve the "shared" attribute from a Triton binary (#120666)

chenyang78 committed 207 days ago

c29ac05a

Decomposition for upsample_linear{1d, 3d} (#114774)

isuruf committed 207 days ago

435063aa

Fix test failure: Add torch.cuda._get_device_properties to dynamo trace rules (#120620)

kadeng committed 207 days ago

2ad66e6b

[dynamo] Desugar accumulate_grad, fix .grad handling (#120590)

anijain2305 committed 207 days ago

e3d64c4d

[FSDP] Clean missing and unexpected keys (#120600)

rohan-varma committed 207 days ago

9db6a849

[PyTorch][ExportedProgram] add 'is_lifted_tensor_constant' and 'get_lifted_tensor_constant' utils (#120546)

mcr229 committed 207 days ago

b2a318d8

Increased compile time max GPUs to 512. Switched to int16_t DeviceIndex. (#119639)

tringwald committed 207 days ago

7c556428

[pytree][reland] Require pytree serialized_type_name (#120636)

angelayi committed 207 days ago

cbbc309c

[export] preserve constant fqn (#120664)

suo committed 207 days ago

12f724c7

Keep test order due to rename_privateuse1_backend is disposable (#120464)

dilililiwhy committed 207 days ago

a358b23a

[BE]: Enable ruff LOG checks (#120674)

Skylion007 committed 207 days ago

5a5b6544

Newer Older