pesi
/

rtmo

rtmo / symbolic_shape_infer_all.sh

Make ONNX models compatible with ONNXruntime's TensorrtExecutionProvider

b75f05d over 2 years ago

135 Bytes

	for f in $(ls -S *.onnx \| tac);
	do
	echo Process "$f"
	python3 symbolic_shape_infer.py --input "$f" --output "$f" --auto_merge;
	done