Closed
Description
Bug Description
I created a TRTorch 0.0.3 for Windows and have noticed it crashes at runtime when registering the operators.
Schema:
trt::execute_engine(Tensor[] inputs, __torch__.torch.classes.tensorrt.Engine engine) -> Tensor[]
Function it dies in, while parsing schema:
https://github.com/pytorch/pytorch/blob/4a3aad354a14edb1337c6214c3d3870c136d5286/torch/csrc/jit/frontend/schema_type_parser.cpp#L307
To Reproduce
Steps to reproduce the behavior:
- Use Windows
- Use my precompiled TRTorch
- Link to TRTorch
Prebuilt is available in PR #152 or for debug library:
trtorch-0.0.3-debug.zip
Expected behavior
Runs
Environment
Build information about the TRTorch compiler can be found by turning on debug messages
- PyTorch Version (e.g., 1.0): 1.6.0-rc3
- CPU Architecture: x64
- OS (e.g., Linux): Windows
- How you installed PyTorch (
conda
,pip
,libtorch
, source): libtorch - Build command you used (if compiling from source):
- Are you using local sources or building from archives:
- Python version: 3.7
- CUDA version: 10.2
- GPU models and configuration: RTX 2060
- Any other relevant information: