strisunshinewentingwang/pytorch

Pull Requests Commits

[inductor] Replaced refs.op by torch.op in _refs/* (#102176)

vfdev-5 committed 1 year ago

319a1cb4

[inductor] fix issue with ops.lookup_seed (#102485)

jansel committed 1 year ago

fc0fed36

[inductor] Handle floordiv and remainder in IndexPropagation (#102277)

peterbell10 committed 1 year ago

c6d9a0b9

[inductor] Inline ComputedBuffer computation when there are no reads (#102000)

peterbell10 committed 1 year ago

e4e151d6

[inductor] erfinv: CPU/CUDA lowering (#101863)

kshitij12345 committed 1 year ago

b1bc8aec

Revert "Replace int64_t with a size type in python_list.h when applicable (#101922)"

pytorchmergebot committed 1 year ago

0803b918

Improve precision and performance for BFloat16 upsampling (#91169)

CaoE committed 1 year ago

af1d4376

[dynamo] Some torchrec_dlrm related fixes (#101953)

anijain2305 committed 1 year ago

040d2cc9

Enable CuDNN v8 frontend in RL (#102284)

malfet committed 1 year ago

53d1d301

Revert "[FSDP]Add device_mesh to FSDPstate (#102317)"

pytorchmergebot committed 1 year ago

81ac076b

[PT2][Quant] Enable test_qnnpack_quantizer_conv_linear test (#102399)

kimishpatel committed 1 year ago

af70fe9f

[PT2][Quant] Move observer sharing ops to use module partitions (#102398)

kimishpatel committed 1 year ago

0d876f7d

[PT2][Quant] Move add/add relu pattern via module partitioner (#102397)

kimishpatel committed 1 year ago

9fac5afb

[PT2][Quant] Move maxpool_2d quant to use module partitioners (#102396)

kimishpatel committed 1 year ago

3d8f4050

[PT2][Quant] Use module partitions for conv2d and conv2d + relu (#102395)

kimishpatel committed 1 year ago

d997e3aa

[PT2][Quant] Use module partition for fused patterns (#102394)

kimishpatel committed 1 year ago

4cb6add4

[FSDP]Add device_mesh to FSDPstate (#102317)

wz337 committed 1 year ago

4c584acc

[pt2] convert `out` params in `register_meta` (#101344)

nkaretnikov committed 1 year ago

c3ea8cc5

Replace int64_t with a size type in python_list.h when applicable (#101922)

pandaninjas committed 1 year ago

44e7f07e

add Half support for logsigmoid, threshold, elu, gelu, hardtanh, hardsigmoid, hardswish, hardshrink, softshrink, leakyrelu, softplus, glu, silu, mish, and prelu on CPU (#98745)

CaoE committed 1 year ago

3f4fee73

[quant][pt2e] Fix propagate_annotation after recent refactors (#102422)

jerryzh168 committed 1 year ago

eda5abf5

Do not collect and skip non-disabled tests when rerunning disabled tests (#102107)

huydhn committed 1 year ago

6e3e3dd4

[pt2] add `SymInt` support for `linalg.pinv` (#102367)

nkaretnikov committed 1 year ago

995ac703

Add security guards to avoid crashes in torch::jit module (#102156)

apach301 committed 1 year ago

c9f4f019

Fix some move warnings by gcc13 (#102353)

cyyever committed 1 year ago

d7eec562

Deallocate workspace on thread exit (#102276)

blawrence-ont committed 1 year ago

26f53bb8

perf hint logging in inductor (#102250)

spectrometerHBH committed 1 year ago

5ee46afc

Modified logging threshold for memory profiling (#102243)

BLOrange-AMD committed 1 year ago

25058d5f

Support dynamo tracing collectives with processgroup arg (#102222)

wconstab committed 1 year ago

e344ff41

add additional stream priority for cuda streams (#101956)

ngimel committed 1 year ago

ecd79b1f

Newer Older