def test_trainer(self):
num_batches = 5
x_train = np.random.randint(0, VOCABULARY_SIZE, [BATCH_SIZE * num_batches, SEQUENCE_LENGTH])
y_train = np.eye(2)[np.ones(BATCH_SIZE * num_batches, dtype=np.int32)]
train_data_iter = tfmodels.data.utils.batch_iter(list(zip(x_train, y_train)), BATCH_SIZE, 1, fill=True, seed=42)
train_data_iter = map(lambda batch: zip(*batch), train_data_iter)
losses = []
with self.graph.as_default(), self.sess.as_default():
rnn = self._build_classifier()
t = RNNClassifierTrainer(rnn)
self.sess.run(tf.initialize_all_variables())
train_iter = t.train_loop(train_data_iter)
for loss, acc, current_step, time_delta in train_iter:
losses = losses + [loss]
self.assertEqual(len(losses), 5)
self.assertEqual(current_step, 5)
def test_trainer(self):
num_batches = 5
x_train = np.random.randint(
0, VOCABULARY_SIZE, [BATCH_SIZE * num_batches, SEQUENCE_LENGTH])
y_train = np.eye(2)[np.ones(BATCH_SIZE * num_batches, dtype=np.int32)]
train_data_iter = tfmodels.data.utils.batch_iter(list(
zip(x_train, y_train)),
BATCH_SIZE,
1,
fill=True,
seed=42)
train_data_iter = map(lambda batch: zip(*batch), train_data_iter)
losses = []
with self.graph.as_default(), self.sess.as_default():
rnn = self._build_classifier()
t = RNNClassifierTrainer(rnn)
self.sess.run(tf.initialize_all_variables())
train_iter = t.train_loop(train_data_iter)
for loss, acc, current_step, time_delta in train_iter:
losses = losses + [loss]
self.assertEqual(len(losses), 5)
self.assertEqual(current_step, 5)
}
model_params.update(FLAGS.__flags)
model = RNNClassifier.from_dict(model_params)
model.print_params()
x = tf.placeholder(tf.int32, [FLAGS.batch_size, SEQUENCE_LENGTH])
y = tf.placeholder(tf.float32, [FLAGS.batch_size, 2])
model.build_graph(x, y)
# Directory for training and dev summaries
timestamp = str(int(time.time()))
rundir = os.path.abspath(os.path.join(os.path.curdir, "runs", timestamp))
train_dir = os.path.join(rundir, "train")
dev_dir = os.path.join(rundir, "dev")
# Build the Trainer/Evaluator
trainer = RNNClassifierTrainer(model, train_summary_dir=train_dir)
evaluator = RNNClassifierEvaluator(model, summary_dir=dev_dir)
# Saving/Checkpointing
checkpoint_dir = os.path.join(rundir, "checkpoints")
if not os.path.exists(checkpoint_dir):
os.makedirs(checkpoint_dir)
checkpoint_file = os.path.join(checkpoint_dir, "model.ckpt")
saver = tf.train.Saver(keep_checkpoint_every_n_hours=2)
# Initialization, optinally load from checkpoint
sess.run(tf.initialize_all_variables())
latest_checkpoint = tf.train.latest_checkpoint(checkpoint_dir)
if latest_checkpoint:
print("Restoring checkpoint from {}".format(latest_checkpoint))
saver.restore(sess, latest_checkpoint)
model_params = {"sequence_length": SEQUENCE_LENGTH, "vocabulary_size": VOCABULARY_SIZE, "num_classes": 2}
model_params.update(FLAGS.__flags)
model = RNNClassifier.from_dict(model_params)
model.print_params()
x = tf.placeholder(tf.int32, [FLAGS.batch_size, SEQUENCE_LENGTH])
y = tf.placeholder(tf.float32, [FLAGS.batch_size, 2])
model.build_graph(x, y)
# Directory for training and dev summaries
timestamp = str(int(time.time()))
rundir = os.path.abspath(os.path.join(os.path.curdir, "runs", timestamp))
train_dir = os.path.join(rundir, "train")
dev_dir = os.path.join(rundir, "dev")
# Build the Trainer/Evaluator
trainer = RNNClassifierTrainer(model, train_summary_dir=train_dir)
evaluator = RNNClassifierEvaluator(model, summary_dir=dev_dir)
# Saving/Checkpointing
checkpoint_dir = os.path.join(rundir, "checkpoints")
if not os.path.exists(checkpoint_dir):
os.makedirs(checkpoint_dir)
checkpoint_file = os.path.join(checkpoint_dir, "model.ckpt")
saver = tf.train.Saver(keep_checkpoint_every_n_hours=2)
# Initialization, optinally load from checkpoint
sess.run(tf.initialize_all_variables())
latest_checkpoint = tf.train.latest_checkpoint(checkpoint_dir)
if latest_checkpoint:
print("Restoring checkpoint from {}".format(latest_checkpoint))
saver.restore(sess, latest_checkpoint)
