strisunshinewentingwang/pytorch

Pull Requests Commits

[pytree] Allow register_pytree_node to take in 5 inputs (#108256)

angelayi committed 1 year ago

58268137

[vmap] symintify alias and squeeze (#107577)

kshitij12345 committed 1 year ago

50fa5880

[export] Fix torch.export() issues for server use cases. (#108275)

zhxchen17 committed 1 year ago

138fafe7

Use ctypes to serialize raw content for tensors. (#108287)

muchulee8 committed 1 year ago

43f28bef

clang8=>clang9 in jobs (#107144)

cyyever committed 1 year ago

c24d0d31

[4/N] fix clang-tidy warnings in torch/csrc (#108305)

cyyever committed 1 year ago

a20fac89

[ONNX] Move large scale models without non-persistent buffers to runtime test (#108084)

titaiwangms committed 1 year ago

d72b990b

[release_note_tool] Update test and skip commits that errors out (#108252)

jerryzh168 committed 1 year ago

9ed0b3fc

[Dynamo] SetVariable supports contains (#108189)

yanboliang committed 1 year ago

9862c719

Fallback to eager for float8 ops in inductor (#108293)

drisspg committed 1 year ago

98aa3745

Allow registering decomps for HigherOrderOp; add decomp for out_dtype (#108080)

zou3519 committed 1 year ago

0e4752ba

Revert "[BE] Pin scipy to 1.10.1 (#108270)"

pytorchmergebot committed 1 year ago

95e31263

Added info for each artifact option, added a help option to TORCH_LOGS, and changed the error message (#107758)

Chillee committed 1 year ago

11860d9d

[BE] Pin scipy to 1.10.1 (#108270)

malfet committed 1 year ago

cd3860cf

Update to RNN documentation (issue #106085) (#106222)

hasteinmetz committed 1 year ago

b535ed2c

Fix triton upload channel detection (#108291)

huydhn committed 1 year ago

23a6706c

[inductor] no-side-effect codegen (#107617)

shunting314 committed 1 year ago

7cb4bf67

Fix layernorm cpu precision issues (#108089)

mikaylagawarecki committed 1 year ago

3817de5d

[inductor] Fix MKL issue with test_indirect_device_assert (#108172)

jansel committed 1 year ago

8a089f63

[inductor] Ignore sympy.PolynomialError while simplifying (#108280)

jansel committed 1 year ago

b2fe5eb7

[inductor] Move test_inductor_sequence_nr out of test_aot_inductor (#108237)

desertfire committed 1 year ago

68304809

[memory snapshots] _record_memory_history_legacy bug fix (#108260)

zdevito committed 1 year ago

7fb13104

Horizontally fuse input concatenation (#108115)

mlazos committed 1 year ago

5911faeb

[RESUBMIT] Standardize on error types for distributed errors. (#108191)

pritamdamania87 committed 1 year ago

704b0b3c

[submodule] [C10] Update gloo. (#107236)

Rodrigo Kumpera committed 1 year ago

6dacf52f

Add reinplacing pass for scatters + incremental fake tensor updating (#106192)

Chillee committed 1 year ago

39130c74

reduce overhead in split and chunk for NestedTensor (#108213)

cpuhrsch committed 1 year ago

d0b725ea

[inductor] Add input generation fn option for autotuning (#108242)

aakhundov committed 1 year ago

071f9ccd

Faster gc_count update for CUDACachingAllocator (#108071)

doehyun committed 1 year ago

ad17e5ec

[TD] Emit metrics to compare heuristic quality (#108192)

ZainRizvi committed 1 year ago

238cc84a

Newer Older