-
Notifications
You must be signed in to change notification settings - Fork 4
/
model_train.py
176 lines (138 loc) · 7.62 KB
/
model_train.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
import tensorflow as tf
import numpy as np
import os
import time
import datetime
from StanfordReader import StanfordReader
from tensorflow.contrib import learn
import data_utils
import pickle
import sys
data_path = "/Users/kellyzhang/Documents/ReadingComprehension/reading-comprehension/deploy/data/"
# ======================== MODEL HYPERPARAMETERS ========================================
tf.flags.DEFINE_float("dropout_keep_prob", 0.5, "Dropout keep probability (default: 0.5)")
tf.flags.DEFINE_float("learning_rate", 0.001, "Learning rate")
tf.flags.DEFINE_float("l2_reg_lambda", 0.0, "Weight lambda on l2 regularization")
# Training Parameters
tf.flags.DEFINE_integer("batch_size", 64, "Batch Size")
tf.flags.DEFINE_integer("num_epochs", 20, "Number of training epochs (default: 200)")
tf.flags.DEFINE_boolean("shuffle", True, "Shuffle between batches boolean")
# Display/Saving Parameters
tf.flags.DEFINE_integer("print_every", 10, "Print train step after this many steps (default: 100)")
tf.flags.DEFINE_integer("evaluate_every", 10, "Evaluate model on dev set after this many steps (default: 100)")
tf.flags.DEFINE_integer("checkpoint_every", 10, "Save model after this many steps (default: 100)")
# Misc Parameters
tf.flags.DEFINE_boolean("allow_soft_placement", True, "Allow device soft device placement")
tf.flags.DEFINE_boolean("log_device_placement", False, "Log placement of ops on devices")
# Print
FLAGS = tf.flags.FLAGS
FLAGS._parse_flags()
print("\nParameters:")
for attr, value in sorted(FLAGS.__flags.items()):
print("{}={}".format(attr.upper(), value))
print("")
# =============================== PREPARING DATA FOR TRAINING/VALIDATION/TESTING ===============================================
batches = data_utils.create_batches_wdw(num_epochs=FLAGS.num_epochs, batch_size=FLAGS.batch_size,\
shuffle=FLAGS.shuffle, data_path=data_path, dataset="train", old=True, num_examples=100, vocab_size=50000)
dev_batch = data_utils.create_batches_wdw(num_epochs=1, batch_size=10000, \
shuffle=False, data_path=data_path, dataset="val", old=True, num_examples=1000, vocab_size=50000)
for batch in dev_batch:
d_padded_val, q_padded_val, c_indices_val, a_indices_val = data_utils.pad_batch_wdw(batch, train=False)
# ================================================== MODEL TRAINING ======================================
with tf.Graph().as_default():
session_conf = tf.ConfigProto(
allow_soft_placement=FLAGS.allow_soft_placement,
log_device_placement=FLAGS.log_device_placement
)
session_conf.gpu_options.allow_growth = True
sess = tf.Session(config=session_conf)
with sess.as_default():
stan_reader = StanfordReader(max_entities=5, batch_size=FLAGS.batch_size)
# Define Training procedure
global_step = tf.Variable(0, name="global_step", trainable=False)
optimizer = tf.train.AdamOptimizer(learning_rate = FLAGS.learning_rate)
# aggregation_method is an experimental feature introduced for faster gradient computation
grads_and_vars = optimizer.compute_gradients(stan_reader.loss, aggregation_method = 2)
clipped_grads = []
for g, v in grads_and_vars:
if g is not None:
clipped = tf.clip_by_norm(g, clip_norm=10.)
clipped_grads.append((clipped, v))
train_op = optimizer.apply_gradients(clipped_grads, global_step=global_step)
# Keep track of gradient values and sparsity (optional)
grad_summaries = []
for g, v in grads_and_vars:
if g is not None:
grad_hist_summary = tf.histogram_summary("{}/grad/hist".format(v.name), g)
sparsity_summary = tf.scalar_summary("{}/grad/sparsity".format(v.name), tf.nn.zero_fraction(g))
grad_summaries.append(grad_hist_summary)
grad_summaries.append(sparsity_summary)
grad_summaries_merged = tf.merge_summary(grad_summaries)
# Output directory for models and summaries
timestamp = str(int(time.time()))
out_dir = os.path.abspath(os.path.join(os.path.curdir, "runs", timestamp))
print("Writing to {}\n".format(out_dir))
# Summaries for loss and accuracy
loss_summary = tf.scalar_summary("loss", stan_reader.loss)
acc_summary = tf.scalar_summary("accuracy", stan_reader.accuracy)
# Train Summaries
train_summary_op = tf.merge_summary([loss_summary, acc_summary, grad_summaries_merged])
train_summary_dir = os.path.join(out_dir, "summaries", "train")
train_summary_writer = tf.train.SummaryWriter(train_summary_dir, sess.graph)
# Dev summaries
dev_summary_op = tf.merge_summary([loss_summary, acc_summary])
dev_summary_dir = os.path.join(out_dir, "summaries", "dev")
dev_summary_writer = tf.train.SummaryWriter(dev_summary_dir, sess.graph)
# Checkpoint directory. Tensorflow assumes this directory already exists so we need to create it
checkpoint_dir = os.path.abspath(os.path.join(out_dir, "checkpoints"))
checkpoint_prefix = os.path.join(checkpoint_dir, "model")
if not os.path.exists(checkpoint_dir):
os.makedirs(checkpoint_dir)
saver = tf.train.Saver(tf.all_variables())
# Initialize all variables
sess.run(tf.initialize_all_variables())
def train_step(train_d, train_q, train_choices, train_answer, print_bool=True):
#A single training step
feed_dict = {
stan_reader.input_d : train_d,
stan_reader.input_q : train_q,
stan_reader.input_a : train_choices,
stan_reader.input_m : train_answer
}
_, step, summaries, loss, accuracy = sess.run(
[train_op, global_step, train_summary_op, stan_reader.loss, stan_reader.accuracy],
feed_dict)
if print_bool:
time_str = datetime.datetime.now().isoformat()
print("{}: step {}, loss {:g}, acc {:g}".format(time_str, step, loss, accuracy))
train_summary_writer.add_summary(summaries, step)
def dev_step(val_d, val_q, val_choices, val_answer, writer=None):
# Evaluates model on a dev set
feed_dict = {
stan_reader.input_d : val_d,
stan_reader.input_q : val_q,
stan_reader.input_a : val_choices,
stan_reader.input_m : val_answer
}
step, summaries, loss, accuracy = sess.run(
[global_step, dev_summary_op, stan_reader.loss, stan_reader.accuracy],
feed_dict)
time_str = datetime.datetime.now().isoformat()
print("{}: step {}, loss {:g}, acc {:g}".format(time_str, step, loss, accuracy))
if writer:
writer.add_summary(summaries, step)
# Generate batches
for batch in batches:
current_step = tf.train.global_step(sess, global_step)
if current_step % FLAGS.print_every == 0:
print_bool = True
else:
print_bool = False
d_padded, q_padded, c_indices, a_indices = data_utils.pad_batch_wdw(batch, train=True)
train_step(d_padded, q_padded, c_indices, a_indices, print_bool=print_bool)
if current_step % FLAGS.evaluate_every == 0:
print("\nEvaluation:")
#dev_step(d_padded_val, q_padded_val, c_indices_val, a_indices_val, writer=dev_summary_writer)
if current_step % FLAGS.checkpoint_every == 0:
path = saver.save(sess, checkpoint_prefix, global_step=current_step)
print("Saved model checkpoint to {}\n".format(path))