onnx2tf
onnx2tf converts ONNX model files to LiteRT, TensorFlow Lite, TensorFlow, native PyTorch modules, TorchScript, PyTorch state dictionaries and exported PyTorch programs, and can also convert directly from LiteRT back to PyTorch.
onnx2tf is a command-line tool for converting ONNX model files to LiteRT, TensorFlow Lite, TensorFlow, native PyTorch code as an nn.Module, TorchScript, a PyTorch state dictionary, and an Exported Program, as well as Dynamo ONNX. It also supports converting directly from LiteRT back to PyTorch.
Its internal tf_converter component tracks per-operator support status against the ONNX operator specification, marking each op as fully supported, partially supported, or open for a contribution, covering the great majority of standard ONNX operators including attention, convolution, quantization and Fourier-transform layers. The project notes that Google's own LiteRT Torch tool is now the recommended path for that particular conversion direction, and points to it and a companion AI Edge quantizer tool as alternatives.