def get_config(config_size):
config_size = config_size.lower()
if config_size == 'small':
return c.SmallConfig()
elif config_size == 'medium':
return c.MediumConfig()
elif config_size == 'large':
return c.LargeConfig()
else:
raise ValueError('Unknown config size {} (small, medium, large)'.format(config_size))
def DataTests():
model_config = config.MediumConfig()
train_data = PTB_DATA()
train_data.load_data(config.TRAIN_FILENAME, model_config.batch_size)
total_batch = int((train_data.batch_len - 1)/model_config.num_steps)
shape = (model_config.batch_size, model_config.num_steps)
for epoch in range(2):
for batch in range(total_batch):
batch_X, batch_Y = train_data.generate_batch(model_config.batch_size, model_config.num_steps, batch)
print("batch:", batch)
assert batch_X.shape == shape
assert batch_Y.shape == shape
@author: rahul.ghosh
"""
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "1"
import data
import model
import config
import tensorflow as tf
if __name__ == "__main__":
with tf.Graph().as_default():
# LOAD CONFIG
model_config = config.MediumConfig()
eval_config = config.MediumConfig()
gen_config = config.SmallGenConfig()
eval_config.batch_size = 1
eval_config.num_steps = 1
# READ DATA
train_data = data.PTB_DATA()
train_data.load_data(config.TRAIN_FILENAME, model_config.batch_size)
# valid_data = data.PTB_DATA()
# valid_data.load_data(config.VALIDATION_FILENAME, model_config.batch_size)
test_data = data.PTB_DATA()
test_data.load_data(config.TEST_FILENAME, eval_config.batch_size)
# BUILD MODEL
initializer = tf.random_uniform_initializer(-model_config.init_scale,
model_config.init_scale)
with tf.name_scope("Train"):