Convert TF SavedModel To ONNX

Public technical reference documentation from Graiphic. For current SOTA installation and toolkit guides, visit the documentation center.

Description

This VI converts a TensorFlow model saved in the SavedModel directory format into a .onnx file, using the tf2onnx converter. It enables TensorFlow-trained models to be deployed in ONNX-compatible runtimes and platforms.
Documentation illustration (illustration unavailable in the archive).

Input parameters

Open Netron : boolean, indicating whether to automatically open the resulting ONNX file in Netron after conversion. If true opens in Netron else conversion only.
TF Model Folder Path : path, path to the TensorFlow SavedModel folder. This directory must contain saved_model.pb and the associated variables/subfolder.
ONNX Model Path : path, destination file path for the exported .onnx model.

Output parameters

standard output : string, text output from the underlying Python process. Can include logs, conversion info, or warnings.
standard error : string, text output capturing any error messages from the Python process, useful for debugging failed conversions.

Example

All these exemples are snippets PNG, you can drop these Snippet onto the block diagram and get the depicted code added to your VI (Do not forget to install Deep Learning library to run it).