81 lines
2.8 KiB
Python
81 lines
2.8 KiB
Python
# Copyright 2020 The TensorFlow Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
# ==============================================================================
|
|
"""Python utility script to generate unit test model data."""
|
|
|
|
# Steps to regenerate model test data:
|
|
# TODO(b/158011574): Do these steps in the script here instead of manually.
|
|
# 1.) Run this script
|
|
# 2.) Hexdump the model into a .h/.cc file:
|
|
# xxd -i /tmp/tf_micro_conv_test_model.tflite > /tmp/temp.cc
|
|
# 3.) Copy/replace contents of temp.cc into desired header/source files (e.g.
|
|
# test_conv_model.h/.cc
|
|
|
|
from __future__ import absolute_import
|
|
from __future__ import division
|
|
from __future__ import print_function
|
|
|
|
from absl import app
|
|
import numpy as np
|
|
import tensorflow.compat.v2 as tf
|
|
|
|
|
|
def generate_conv_model():
|
|
"""Creates a basic Keras model and converts to tflite.
|
|
|
|
This model does not make any relevant classifications. It only exists to
|
|
generate a model that is designed to run on embedded devices.
|
|
"""
|
|
input_shape = (16, 16, 1)
|
|
|
|
model = tf.keras.models.Sequential()
|
|
model.add(
|
|
tf.keras.layers.Conv2D(16, 3, activation="relu", input_shape=input_shape))
|
|
model.add(tf.keras.layers.Conv2D(32, 3, activation="relu"))
|
|
model.add(tf.keras.layers.MaxPooling2D(2))
|
|
model.add(tf.keras.layers.Flatten())
|
|
model.add(tf.keras.layers.Dense(10))
|
|
model.compile(
|
|
optimizer="adam", loss="categorical_crossentropy", metrics=["accuracy"])
|
|
model.summary()
|
|
|
|
# Test with random data
|
|
data_x = np.random.rand(12, 16, 16, 1)
|
|
data_y = np.random.randint(2, size=(12, 10))
|
|
model.fit(data_x, data_y, epochs=5)
|
|
|
|
def representative_dataset_gen():
|
|
for _ in range(12):
|
|
yield [np.random.rand(16, 16).reshape(1, 16, 16, 1).astype(np.float32)]
|
|
|
|
# Now convert to a TFLite model with full int8 quantization:
|
|
converter = tf.lite.TFLiteConverter.from_keras_model(model)
|
|
converter.optimizations = [tf.lite.Optimize.DEFAULT]
|
|
converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8]
|
|
converter.inference_input_type = tf.int8
|
|
converter.inference_output_type = tf.int8
|
|
converter.representative_dataset = representative_dataset_gen
|
|
|
|
tflite_model = converter.convert()
|
|
open("/tmp/tf_micro_conv_test_model.int8.tflite", "wb").write(tflite_model)
|
|
|
|
|
|
def main(argv):
|
|
del argv # Unused for now
|
|
generate_conv_model()
|
|
|
|
|
|
if __name__ == "__main__":
|
|
app.run(main)
|