Skip to content

ModelOpt 0.46.1 Release

Latest

Choose a tag to compare

@kevalmorabia97 kevalmorabia97 released this 09 Sep 16:33
b5c76a0

WoA (Windows) Support

Add opt-in TensorRT-RTX ABI Execution Provider support for ONNX calibration on Windows arm64. Select it with --calibration_eps=NvTensorRtRtx --trt_rtx_backend=abi; the legacy backend remains the default.