train.py

#  MIT License
#
#  Copyright (c) 2020 Omer Ferhat Sarioglu
#
#  Permission is hereby granted, free of charge, to any person obtaining a copy
#  of this software and associated documentation files (the "Software"), to deal
#  in the Software without restriction, including without limitation the rights
#  to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
#  copies of the Software, and to permit persons to whom the Software is
#  furnished to do so, subject to the following conditions:
#
#  The above copyright notice and this permission notice shall be included in all
#  copies or substantial portions of the Software.
#
#  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
#  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
#  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
#  AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
#  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
#  OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
#  SOFTWARE.

import tensorflow as tf
import argparse

from model.model import create_model, LearningRateScheduler
from data.data_generator import DatasetGenerator


argp = argparse.ArgumentParser()
argp.add_argument("-r", "--raw-path", type=str, default="data/raw",
                  help="Raw signal path.")
argp.add_argument("-a", "--annot-path", type=str, default="data/annotations/csv",
                  help="Path of signal annotations.")
argp.add_argument("-b", "--batch-size", type=int, default=128,
                  help="Number of batch size.")
argp.add_argument("-e", "--epoch", type=int, default=10,
                  help="Number of epoch.")
argp.add_argument("-B", "--beat-width", type=int, default=128,
                  help="Sample number of one beat.")
argp.add_argument("-R", "--random-seed", type=int, default=5,
                  help="Enter `0` for non-random arrays.")
argp.add_argument("-l", "--log-dir", type=str, default="model/logs",
                  help="Folder to save TensorBoard files.")
argp.add_argument("-m", "--model-file", type=str, default="model",
                  help="Name of the *.h5 file.")

args = argp.parse_args()


BEAT_WIDTH = args.beat_width
EPOCH = args.epoch
BATCH_SIZE = args.batch_size

model = create_model(beat_width=BEAT_WIDTH)
model.summary()

data_generator = DatasetGenerator(raw_path=args.raw_path,
                                  annot_path=args.annot_path,
                                  beat_width=BEAT_WIDTH,
                                  random_seed=args.random_seed)

tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=args.log_dir, histogram_freq=1)

LR_SCHEDULE = [
    # (epoch to start, learning rate) tuples
    (4, 0.001), (8, 0.0005), (12, 0.0002), (15, 0.0001)
]


def lr_schedule(epoch, lr):
    """Helper function to retrieve the scheduled learning rate based on epoch."""
    if epoch < LR_SCHEDULE[0][0] or epoch > LR_SCHEDULE[-1][0]:
        return lr
    for i in range(len(LR_SCHEDULE)):
        if epoch == LR_SCHEDULE[i][0]:
            return LR_SCHEDULE[i][1]
    return lr


model.fit(data_generator.X_train, data_generator.y_train,
          epochs=EPOCH, batch_size=BATCH_SIZE, validation_split=0.1,
          callbacks=[tensorboard_callback, LearningRateScheduler(lr_schedule)])

model.save("model/logs/{}.h5".format(args.model_file))

loss, acc = model.evaluate(data_generator.X_test, data_generator.y_test, batch_size=1)

print("loss={}".format(loss))
print("acc=%{}".format(acc))