Convert tensorflow-hub model to tensorflow lite(tflite)-CodePudding

I was trying to convert BigGAN model in tensorflow-hub(.pb) to a TensorFlow Lite file (.tflite) using the following code:

import tensorflow.compat.v1 as tf
tf.disable_v2_behavior()
import numpy as np
import tensorflow_hub as hub

module_path = 'https://tfhub.dev/deepmind/biggan-deep-256/1'
tf.compat.v1.reset_default_graph()
print('Loading BigGAN module from:', module_path)
module = hub.Module(module_path)
dummy_inputs = {
    "y": tf.compat.v1.placeholder(tf.float32, [1, 1000], 'y'),
    "z": tf.compat.v1.placeholder(tf.float32, [1, 128], 'z'),
    "truncation": tf.compat.v1.placeholder(tf.float32, [], 'truncation'),
}
dummy_output = module(dummy_inputs)
print('dummy_Inputs:\n', '\n'.join(
    '  {}: {}'.format(*kv) for kv in dummy_inputs.items()))
print('dummy_Output:', dummy_output)

initializer = tf.global_variables_initializer()
sess = tf.Session()
sess.run(initializer)

save_path = "./big_gan.tflite"
_input = [dummy_inputs[k] for k in dummy_inputs]
converter = tf.lite.TFLiteConverter.from_session(sess, _input, [dummy_output])
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.experimental_new_converter = True

# errors here
tflite_model = converter.convert()

open(save_path, "wb").write(tflite_model)

returned infomations:

2022-01-28 17:10:30.240145: E tensorflow/core/grappler/grappler_item_builder.cc:670] Init node AssignVariableOp_1003 doesn't exist in graph
2022-01-28 17:10:33.853842: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:363] Ignored output_format.
2022-01-28 17:10:33.853899: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:366] Ignored drop_control_dependency.

throw a error:

ConverterError: Input 0 of node module_apply_default/cond/AssignVariableOp was passed float from module_apply_default/cond/AssignVariableOp/Switch:0 incompatible with expected resource.

The module is working as expected follow by "Example use", it can generate images normally.
Can anybody help me to understand what does Init node AssignVariableOp_1003 doesn't exist in graph mean and how to fix the error?

CodePudding user response：

Using TF 2.x to convert a TF 1.x model to a TensorFlow Lite file is tricky. I would recommend running your code example on Google Colab and switching to TF 1.x:

%tensorflow_version 1.x

It seems to work.