How to find out version of Tensorflow used in HDF5 model

Question

After I have saved a tensorflow model in HDF5 format, how can I find out what version of tensorflow it was built using?

Answer 1

The only thing I am aware of, when loading a model with the Keras h5 format, is using model.to_json(), but you will only get the Keras version and backend that were used:

import tensorflow as tf
import numpy as np
import json

model = tf.keras.Sequential([
  tf.keras.layers.Input((256, 256, 3)),
  tf.keras.layers.Conv2D(16, 3, padding='same', activation='relu', strides=1, 
     data_format='channels_last'),
  tf.keras.layers.MaxPooling2D(),
  tf.keras.layers.Flatten(),
  tf.keras.layers.Dense(128, activation='relu'),
  tf.keras.layers.Dropout(0.1),
  tf.keras.layers.Dense(5, activation = 'softmax')
 ])

model.compile(loss='categorical_crossentropy', optimizer='adam')
model.save("test.h5", save_format='h5')

model = tf.keras.models.load_model('test.h5')
meta_data = json.loads(model.to_json())
print('keras_version:', meta_data['keras_version'], 'backend:', meta_data['backend'])

keras_version: 2.7.0 backend: tensorflow

So you would have to find out which Tensorflow versions are compatible with this Keras version, but should not be too difficult.