def main(): args = params.getArgs() x_train, x_test = get_data(args) modelDict = model_IO.load_autoencoder(args) encoder = modelDict.encoder generator = modelDict.generator indices = np.random.choice(len(x_train), 1000, replace=False) images = x_train[indices, :] z_sampled, z_mean, z_logvar = encode(encoder, images, args.batch_size) visualize(z_sampled, z_mean, z_logvar)
from scipy.sparse import dok_matrix import time from itertools import accumulate from scipy.stats import entropy import params from util import get_max_fea # divide GPUs, by randomly selecting one for each process gpus = os.environ["CUDA_VISIBLE_DEVICES"].split(',') gpu_count = len(gpus) gpu = gpus[np.random.randint(0, gpu_count)] os.environ["CUDA_VISIBLE_DEVICES"] = gpu # load parameters args = params.getArgs() # print("\n\n") # for k in args.keys(): # print(k, ": ", args[k]) # print("\n\n") if args.model_type in ("gnn", "nn"): import tensorflow as tf print("Num GPUs Available: ", len(tf.config.experimental.list_physical_devices('GPU'))) gpus = tf.config.experimental.list_physical_devices('GPU') if gpus: try: for gpu in gpus: tf.config.experimental.set_memory_growth(gpu, True) except RuntimeError as e:
model.load_weights('model.h5') wavs = model.predict(wavs) wavs = wavs / config.multiplier wavs = tf.reshape(wavs, [*wavs.shape[:2], 3, 10]) angles = tf.cast(tf.round(tf.reduce_sum(wavs, axis=(1, 2))), tf.int8) classes = tf.cast(tf.round(tf.reduce_sum(wavs, axis=(1, 3))), tf.int8) # import numpy as np # from glob import glob # num = len(sorted(glob('dataset/3rd_track3/*.wav'))) # angles = tf.convert_to_tensor(np.zeros((num, 10), dtype=np.int)) # classes = tf.convert_to_tensor(np.ones((num, 3), dtype=np.int) * 2) data = {'track3_results': list()} for idx, (ag, cl) in enumerate(zip(angles, classes)): _data = { 'id': idx, 'angle': ag.numpy().tolist(), 'class': cl.numpy().tolist() } data['track3_results'].append(_data) tojson(data) if __name__ == "__main__": import sys main(getArgs(sys.argv[1:]))