def choreograph_dance(model, save_location, out_file):
if(not model):
#loads the "best" model
model = generator.load_trained_model(architecture_file, logs_save_dir)
model = generator.compile_model(model, optimizer, loss_function, metrics)
utils.write(generator.model_summary(model), out_file)
print(generator.model_summary(model))
generator.generate_dance(model, frames, out_file = out_file)
def evaluate_generator(model, eval_X, eval_Y, save_location, out_file):
if(not model):
#loads the "best" model
model = generator.load_trained_model(architecture_file, logs_save_dir)
model = generator.compile_model(model, optimizer, loss_function, metrics)
utils.write(generator.model_summary(model), out_file)
print(generator.model_summary(model))
eval_X = np.load(evaluation_filepath+"_X.npy")
eval_Y = np.load(evaluation_filepath+"_Y.npy")
return generator.benchmark(model, eval_X, eval_Y, out_file)
def train_generator(out_file=sys.stdout):
""" Trains the model with the dance data.
The History object's History.history attribute is a record of training loss values and metrics values at successive epochs,
as well as cooresponding validation values (if applicable).
:param out_file: what to display/write the status information to
:type output stream
:return: the class containing the training metric information, the trained model, and the comprehensive evaluation data
:type tuple
"""
model = generator.establish_model(units, look_back, n_features, layer_activation, recurrent_activation, weight_initializer, recurrent_initializer, bias_initializer, output_activation, mixtures, use_mdn)
generator.save_architecture(model, architecture_file)
model = generator.compile_model(model, optimizer, loss_function, metrics)
utils.write(generator.model_summary(model), out_file)
print(generator.model_summary(model))
return generator.train_model(model, out_file)
def choreograph_dance(model, save_location, out_file):
if (not model):
#loads the most recent saved model
filename = [f for f in os.listdir(logs_save_dir) if "model" in f][-1]
model = generator.load_trained_model(
os.path.join(save_location, filename))
utils.write(generator.model_summary(model), out_file)
generator.generate_dance(model, frames, out_file=out_file)
def evaluate_generator(model, save_location, out_file):
if (not model):
#loads the most recent saved model
filename = [f for f in os.listdir(logs_save_dir) if "model" in f][-1]
model = generator.load_trained_model(
os.path.join(save_location, filename))
utils.write(generator.model_summary(model), out_file)
eval_X = np.load(
os.path.join(
np_save_dir, "_comprehensive_evaluation_" + data_identifier +
"_es-{}".format(evaluation_split) + "_X.npy"))
eval_Y = np.load(
os.path.join(
np_save_dir, "_comprehensive_evaluation_" + data_identifier +
"_es-{}".format(evaluation_split) + "_Y.npy"))
return generator.benchmark(model, eval_X, eval_Y, out_file)