frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-21 05:34:18 +08:00
Code Issues Packages Projects Releases Wiki Activity

Revert D24924236: [pytorch][PR] [ONNX] Handle sequence output shape and type inference

Browse Source 
Test Plan: revert-hammer

Differential Revision:
D24924236 (adc65e7c8d)

Original commit changeset: 506e70a38cfe

fbshipit-source-id: 78069a33fb3df825af1cb482da06a07f7b26ab48
This commit is contained in:
Brian Vaughan
2021-01-15 05:56:51 -08:00
committed by Facebook GitHub Bot
parent 366b00ab7b
commit a9db2f8e7a
20 changed files with 71 additions and 248 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
torch/csrc/jit/python/init.cpp
View File

@ -150,9 +150,8 @@ void initJITBindings(PyObject* module) {
"_jit_pass_onnx_assign_output_shape",
[](std::shared_ptr<Graph>& graph,
const std::vector<at::Tensor>& tensors,
const python::IODescriptor& desc,
bool onnx_shape_inference = false) {
ONNXAssignOutputShape(graph, tensors, desc, onnx_shape_inference);
ONNXAssignOutputShape(graph, tensors, onnx_shape_inference);
})
.def("_jit_pass_lower_all_tuples", LowerAllTuples)
.def("_jit_pass_onnx_function_substitution", ONNXFunctionCallSubstitution)