TensorRT
TensorRT copied to clipboard
✨[Feature] ONNX vs. ATen Perf Testing.
Is your feature request related to a problem? Please describe.
Compare ONNX vs ATen subgraph perf
Describe the solution you'd like
Describe alternatives you've considered
Additional context
Beyond https://github.com/pytorch/TensorRT/issues/3733 focused on the num of ILayer types, we need to benchmark latency as well.