Tensorflow: What are the "output_node_names" for freeze_graph.py in the model_with_buckets model?

The "output_node_names" for the "freeze_graph.py" script in the "model_with_buckets" model in TensorFlow depend on the specific model architecture and the names assigned to the output nodes. To determine the "output_node_names," you need to inspect the TensorFlow graph of the model.

Here are the general steps to find the "output_node_names":

1. Load the TensorFlow model: Load the saved TensorFlow model using the appropriate method, such as tf.saved_model.load(), tf.keras.models.load_model(), or tf.compat.v1.train.import_meta_graph().

2. Inspect the TensorFlow graph: Once the model is loaded, you can access the TensorFlow graph using the appropriate method, such as model.graph or tf.compat.v1.get_default_graph().

3. Find the output nodes: In the TensorFlow graph, the output nodes represent the final outputs of the model. These nodes are typically named based on the specific model architecture. You can use methods like graph.get_operations() or graph.get_tensor_by_name() to explore the graph and find the names of the output nodes.

4. Determine the "output_node_names": Once you have identified the names of the output nodes, you can specify them as a comma-separated string in the "output_node_names" parameter of the "freeze_graph.py" script.

Please note that the exact steps may vary depending on the specific model architecture and how the model was saved. It is important to refer to the documentation or code of the specific model implementation to find the appropriate steps for your model.

[[1]]