Torch-TensorRT 2.4.0 introduces C++ runtime support on Windows, enhanced dynamic shape support, and broadens ATen core operators coverage. The release supports PyTorch 2.4, CUDA 12.4, and Python 3.12. Notable features include near 100% converter support for ATen core operators and detailed fall back information using 'dryrun'.
Table of contents
C++ runtime support in Windows Support, Enhanced Dynamic Shape support in Converters, PyTorch 2.4, CUDA 12.4, TensorRT 10.1, Python 3.12What's ChangedNew ContributorsSort: