def testPtbRawData(self):
tmpdir = tf.test.get_temp_dir()
for suffix in "train", "valid", "test":
filename = os.path.join(tmpdir, "ptb.%s.txt" % suffix)
with tf.gfile.GFile(filename, "w") as fh:
fh.write(self._string_data)
# Smoke test
output = reader.ptb_raw_data(tmpdir)
self.assertEqual(len(output), 4)
def main():
print("hola mundo")
# import
import tensorflow as tf
with tf.device('/gpu:1'):
current_path = getcwd() + "\Dataset"
train_data, valid_data, test_data, vocabulary = reader.ptb_raw_data(
current_path)
print("el fin")
def main(_):
for keep in conf.keep_probs:
conf.keep_prob = keep
print("keep_prob:" + str(conf.keep_prob))
if not FLAGS.data_path:
raise ValueError("Must set --data_path to data directory")
gpus = [
x.name for x in device_lib.list_local_devices()
if x.device_type == "GPU"
]
if FLAGS.num_gpus > len(gpus):
raise ValueError(
"Your machine has only %d gpus "
"which is less than the requested --num_gpus=%d." %
(len(gpus), FLAGS.num_gpus))
raw_data = reader.ptb_raw_data(FLAGS.data_path)
train_data, valid_data, test_data, _, dict_emb = raw_data
config = get_config()
# FLAGS.save_path = conf.src_path + "/Res/w2v_1w_400d_att_prePro"
eval_config = get_config()
eval_config.batch_size = 1
# eval_config.num_steps = 1
with tf.Graph().as_default() as graph:
# tensorboard图
# train_input = PTBInput(config=config, data=train_data, name="TrainInput")
# m = PTBModel(is_training=True, config=config, input_=train_input, dict_emb=dict_emb)
# file_writer = tf.summary.FileWriter(FLAGS.save_path, graph=graph)
# file_writer.close()
initializer = tf.random_uniform_initializer(
-config.init_scale, config.init_scale)
with tf.name_scope("Train"):
train_input = PTBInput(config=config,
data=train_data,
name="TrainInput")
with tf.variable_scope("Model",
reuse=None,
initializer=initializer):
m = PTBModel(is_training=True,
config=config,
input_=train_input,
dict_emb=dict_emb)
file_writer = tf.summary.FileWriter(FLAGS.save_path,
graph=graph)
file_writer.close()
tf.summary.scalar("Training Loss", m.cost)
tf.summary.scalar("Learning Rate", m.lr)
tf.summary.scalar("Training allCount", m.allCount)
tf.summary.scalar("Training rightCount0", m.rightCountTopK[0])
tf.summary.scalar("Training rightCount1", m.rightCountTopK[1])
tf.summary.scalar("Training rightCount2", m.rightCountTopK[2])
with tf.name_scope("Valid"):
valid_input = PTBInput(config=config,
data=valid_data,
name="ValidInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mvalid = PTBModel(is_training=False,
config=config,
input_=valid_input,
dict_emb=dict_emb)
tf.summary.scalar("Validation Loss", mvalid.cost)
tf.summary.scalar("Validation allCount", mvalid.allCount)
tf.summary.scalar("Validation rightCount0",
mvalid.rightCountTopK[0])
tf.summary.scalar("Validation rightCount1",
mvalid.rightCountTopK[1])
tf.summary.scalar("Validation rightCount2",
mvalid.rightCountTopK[2])
with tf.name_scope("Test"):
test_input = PTBInput(config=eval_config,
data=test_data,
name="TestInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mtest = PTBModel(is_training=False,
config=eval_config,
input_=test_input,
dict_emb=dict_emb)
tf.summary.scalar("Test allCount", mtest.allCount)
tf.summary.scalar("Test rightCount0", mtest.rightCountTopK[0])
tf.summary.scalar("Test rightCount1", mtest.rightCountTopK[1])
tf.summary.scalar("Test rightCount2", mtest.rightCountTopK[2])
models = {"Train": m, "Valid": mvalid, "Test": mtest}
for name, model in models.items():
model.export_ops(name)
metagraph = tf.train.export_meta_graph()
if tf.__version__ < "1.1.0" and FLAGS.num_gpus > 1:
raise ValueError(
"num_gpus > 1 is not supported for TensorFlow versions "
"below 1.1.0")
soft_placement = False
if FLAGS.num_gpus > 1:
soft_placement = True
util.auto_parallel(metagraph, m)
# with tf.Graph().as_default():
# tf.train.import_meta_graph(metagraph)
for model in models.values():
model.import_ops()
sv = tf.train.Supervisor(logdir=FLAGS.save_path,
save_summaries_secs=3)
config_proto = tf.ConfigProto(
allow_soft_placement=soft_placement) #对session进行参数配置
with sv.managed_session(
config=config_proto
) as session: #自动去logdir中找checkpoint，如果没有的话自动初始化
for i in range(config.max_max_epoch):
lr_decay = config.lr_decay**max(i + 1 - config.max_epoch,
0.0)
m.assign_lr(session, config.learning_rate * lr_decay)
print("Epoch: %d Learning rate: %.3f" %
(i + 1, session.run(m.lr)))
train_perplexity, train_rightCountTopK, train_allCount , train_acc \
= run_epoch(session, m, eval_op=m.train_op,verbose=True)
print(
"Epoch: %d Train Perplexity: %.3f train_allCount: %s train_acc: %s rightCountTopK : %s"
% (i + 1, train_perplexity, train_allCount, train_acc,
train_rightCountTopK))
valid_perplexity, val_rightCountTopK, valid_allCount, valid_acc = run_epoch(
session, mvalid)
print(
"Epoch: %d Valid Perplexity: %.3f valid_allCount: %s valid_acc: %s rightCountTopK : %s"
% (i + 1, valid_perplexity, valid_allCount, valid_acc,
val_rightCountTopK))
test_perplexity, test_rightCountTopK, test_allCount, test_acc = run_epoch(
session, mtest)
print(
"Test Perplexity: %.3f test_allCount: %.3f test_acc: %s rightCountTopK : %s"
% (test_perplexity, test_allCount, test_acc,
test_rightCountTopK))
if FLAGS.save_path:
print("Saving model to %s." % FLAGS.save_path)
sv.saver.save(session,
FLAGS.save_path,
global_step=sv.global_step)
print("Finished!")
i = datetime.datetime.now()
print("当前的日期和时间是 %s" % i)
state = vals["final_state"]
costs += cost
iters += model.input.num_steps
if verbose and step % (model.input.epoch_size // 10) == 10:
print("%.3f perplexity:%.3f speed:%.0f wps" %
(step * 1.0 / model.input.epoch_size, np.exp(
costs / iters), iters * model.input.batch_size /
(time.time() - start_time)))
return np.exp(costs / iters)
# 读取解压后的数据
raw_data = reader.ptb_raw_data('simple-examples/data/')
train_data, valid_data, test_data, _ = raw_data
config = SmallConfig()
eval_config = SmallConfig()
eval_config.batch_size = 1
eval_config.num_steps = 1
#创建默认的Graph，创建模型
with tf.Graph().as_default():
initializer = tf.random_uniform_initializer(-config.init_scale,
config.init_scale)
with tf.name_scope("Train"):
train_input = PTBInput(config=config,
data=train_data,
def main(_):
save_path = FLAGS.save_path + '/' + FLAGS.rnn_mode + '_g' + str(
FLAGS.g) + '_lr' + str(FLAGS.lr)
if not FLAGS.data_path:
raise ValueError("Must set --data_path to PTB data directory")
gpus = [
x.name for x in device_lib.list_local_devices()
if x.device_type == "GPU"
]
if FLAGS.num_gpus > len(gpus):
raise ValueError("Your machine has only %d gpus "
"which is less than the requested --num_gpus=%d." %
(len(gpus), FLAGS.num_gpus))
raw_data = reader.ptb_raw_data(FLAGS.data_path)
train_data, valid_data, test_data, _ = raw_data
config = get_config()
eval_config = get_config()
eval_config.batch_size = 1
eval_config.num_steps = config.num_steps # originally is 1
with tf.Graph().as_default():
initializer = tf.random_uniform_initializer(-config.init_scale,
config.init_scale)
with tf.name_scope("Train"):
train_input = PTBInput(config=config,
data=train_data,
name="TrainInput")
with tf.variable_scope("Model",
reuse=None,
initializer=initializer):
m = PTBModel(is_training=True,
config=config,
input_=train_input)
tf.summary.scalar("Training Loss", m.cost)
tf.summary.scalar("Learning Rate", m.lr)
with tf.name_scope("Valid"):
valid_input = PTBInput(config=config,
data=valid_data,
name="ValidInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mvalid = PTBModel(is_training=False,
config=config,
input_=valid_input)
tf.summary.scalar("Validation Loss", mvalid.cost)
with tf.name_scope("Test"):
test_input = PTBInput(config=eval_config,
data=test_data,
name="TestInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mtest = PTBModel(is_training=False,
config=eval_config,
input_=test_input)
tf.summary.scalar("Test Loss", mtest.cost)
models = {"Train": m, "Valid": mvalid, "Test": mtest}
for name, model in models.items():
model.export_ops(name)
metagraph = tf.train.export_meta_graph()
if tf.__version__ < "1.1.0" and FLAGS.num_gpus > 1:
raise ValueError(
"num_gpus > 1 is not supported for TensorFlow versions "
"below 1.1.0")
soft_placement = False
if FLAGS.num_gpus > 1:
soft_placement = True
util.auto_parallel(metagraph, m)
with tf.Graph().as_default():
tf.train.import_meta_graph(metagraph)
for model in models.values():
model.import_ops()
sv = tf.train.Supervisor(logdir=save_path)
config_proto = tf.ConfigProto(allow_soft_placement=soft_placement)
with sv.managed_session(config=config_proto) as session:
# session = tf_debug.LocalCLIDebugWrapperSession(session)
for i in range(config.max_max_epoch):
lr_decay = config.lr_decay**max(i + 1 - config.max_epoch, 0.0)
m.assign_lr(session, config.learning_rate * lr_decay)
print("Epoch: %d Learning rate: %.3f" %
(i + 1, session.run(m.lr)))
train_perplexity = run_epoch(session,
m,
eval_op=m.train_op,
verbose=True)
print("Epoch: %d Train Perplexity: %.3f" %
(i + 1, train_perplexity))
valid_perplexity = run_epoch(session, mvalid)
print("Epoch: %d Valid Perplexity: %.3f" %
(i + 1, valid_perplexity))
test_perplexity = run_epoch(session, mtest)
print("Test Perplexity: %.3f" % test_perplexity)
with open(save_path + "/log.txt", "w+") as myfile:
myfile.write("\nlearning rate: " + str(FLAGS.lr) + '\n')
myfile.write("final Train Perplexity: " +
str(train_perplexity) + '\n')
myfile.write("final Valid Perplexity: " +
str(valid_perplexity) + '\n')
myfile.write("final Test Perplexity: " + str(test_perplexity) +
'\n')
if save_path:
print("Saving model to %s." % save_path)
sv.saver.save(session, save_path, global_step=sv.global_step)
def main(init_scale, learning_rate, max_grad_norm, num_layers, num_steps,
hidden_size, max_epoch, max_max_epoch, keep_prob, lr_decay,
batch_size, vocab_size, use_tanh, data_path, _run):
# if not FLAGS.data_path:
# raise ValueError("Must set --data_path to PTB data directory")
raw_data = reader.ptb_raw_data(data_path)
train_data, valid_data, test_data, _ = raw_data
config = my_config_object()
eval_config = my_config_object()
eval_config.batch_size = 1
eval_config.num_steps = 1
with tf.Graph().as_default():
initializer = tf.random_uniform_initializer(-config.init_scale,
config.init_scale)
with tf.name_scope("Train"):
train_input = PTBInput(config=config,
data=train_data,
name="TrainInput")
with tf.variable_scope("Model",
reuse=None,
initializer=initializer):
m = PTBModel(is_training=True,
config=config,
input_=train_input)
tf.summary.scalar("Training Loss", m.cost)
tf.summary.scalar("Learning Rate", m.lr)
with tf.name_scope("Valid"):
valid_input = PTBInput(config=config,
data=valid_data,
name="ValidInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mvalid = PTBModel(is_training=False,
config=config,
input_=valid_input)
tf.summary.scalar("Validation Loss", mvalid.cost)
with tf.name_scope("Test"):
test_input = PTBInput(config=eval_config,
data=test_data,
name="TestInput")
with tf.variable_scope("Model",
reuse=True,
initializer=initializer):
mtest = PTBModel(is_training=False,
config=eval_config,
input_=test_input)
sv = tf.train.Supervisor(logdir=FLAGS.save_path)
with sv.managed_session() as session:
for i in range(config.max_max_epoch):
lr_decay = config.lr_decay**max(i + 1 - config.max_epoch, 0.0)
m.assign_lr(session, config.learning_rate * lr_decay)
print("Epoch: %d Learning rate: %.3f" %
(i + 1, session.run(m.lr)))
train_perplexity = run_epoch(session,
m,
eval_op=m.train_op,
verbose=True)
print("Epoch: %d Train Perplexity: %.3f" %
(i + 1, train_perplexity))
_run.log_scalar("training.perplexity", train_perplexity, i + 1)
valid_perplexity = run_epoch(session, mvalid)
print("Epoch: %d Valid Perplexity: %.3f" %
(i + 1, valid_perplexity))
_run.log_scalar("validation.perplexity", valid_perplexity,
i + 1)
test_perplexity = run_epoch(session, mtest)
print("Test Perplexity: %.3f" % test_perplexity)
if FLAGS.save_path:
print("Saving model to %s." % FLAGS.save_path)
sv.saver.save(session,
FLAGS.save_path,
global_step=sv.global_step)
return test_perplexity
