Skip to content

GitLab

Unverified Commit 55bf4b80 authored by Hongkun Yu's avatar Hongkun Yu Committed by GitHub
Browse files

Merge branch 'master' into absl

parents 15e0057f 2416dd9c
Hide whitespace changes
Inline Side-by-side
Showing with 0 additions and 795 deletions
tutorials/rnn/ptb/reader_test.py deleted 100644 → 0
View file @ 15e0057f
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for models.tutorials.rnn.ptb.reader."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os.path
import tensorflow as tf
import reader
class PtbReaderTest(tf.test.TestCase):
def setUp(self):
self._string_data = "\n".join(
[" hello there i am",
" rain as day",
" want some cheesy puffs ?"])
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 testPtbProducer(self):
raw_data = [4, 3, 2, 1, 0, 5, 6, 1, 1, 1, 1, 0, 3, 4, 1]
batch_size = 3
num_steps = 2
x, y = reader.ptb_producer(raw_data, batch_size, num_steps)
with self.test_session() as session:
coord = tf.train.Coordinator()
tf.train.start_queue_runners(session, coord=coord)
try:
xval, yval = session.run([x, y])
self.assertAllEqual(xval, [[4, 3], [5, 6], [1, 0]])
self.assertAllEqual(yval, [[3, 2], [6, 1], [0, 3]])
xval, yval = session.run([x, y])
self.assertAllEqual(xval, [[2, 1], [1, 1], [3, 4]])
self.assertAllEqual(yval, [[1, 0], [1, 1], [4, 1]])
finally:
coord.request_stop()
coord.join()
if __name__ == "__main__":
tf.test.main()
tutorials/rnn/ptb/util.py deleted 100644 → 0
View file @ 15e0057f
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Utilities for Grappler autoparallel optimizer."""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import tensorflow as tf
from tensorflow.core.framework import variable_pb2
from tensorflow.core.protobuf import rewriter_config_pb2
FLAGS = tf.flags.FLAGS
def export_state_tuples(state_tuples, name):
for state_tuple in state_tuples:
tf.add_to_collection(name, state_tuple.c)
tf.add_to_collection(name, state_tuple.h)
def import_state_tuples(state_tuples, name, num_replicas):
restored = []
for i in range(len(state_tuples) * num_replicas):
c = tf.get_collection_ref(name)[2 * i + 0]
h = tf.get_collection_ref(name)[2 * i + 1]
restored.append(tf.contrib.rnn.LSTMStateTuple(c, h))
return tuple(restored)
def with_prefix(prefix, name):
"""Adds prefix to name."""
return "/".join((prefix, name))
def with_autoparallel_prefix(replica_id, name):
return with_prefix("AutoParallel-Replica-%d" % replica_id, name)
class UpdateCollection(object):
"""Update collection info in MetaGraphDef for AutoParallel optimizer."""
def __init__(self, metagraph, model):
self._metagraph = metagraph
self.replicate_states(model.initial_state_name)
self.replicate_states(model.final_state_name)
self.update_snapshot_name("variables")
self.update_snapshot_name("trainable_variables")
def update_snapshot_name(self, var_coll_name):
var_list = self._metagraph.collection_def[var_coll_name]
for i, value in enumerate(var_list.bytes_list.value):
var_def = variable_pb2.VariableDef()
var_def.ParseFromString(value)
# Somehow node Model/global_step/read doesn't have any fanout and seems to
# be only used for snapshot; this is different from all other variables.
if var_def.snapshot_name != "Model/global_step/read:0":
var_def.snapshot_name = with_autoparallel_prefix(
0, var_def.snapshot_name)
value = var_def.SerializeToString()
var_list.bytes_list.value[i] = value
def replicate_states(self, state_coll_name):
state_list = self._metagraph.collection_def[state_coll_name]
num_states = len(state_list.node_list.value)
for replica_id in range(1, FLAGS.num_gpus):
for i in range(num_states):
state_list.node_list.value.append(state_list.node_list.value[i])
for replica_id in range(FLAGS.num_gpus):
for i in range(num_states):
index = replica_id * num_states + i
state_list.node_list.value[index] = with_autoparallel_prefix(
replica_id, state_list.node_list.value[index])
def auto_parallel(metagraph, model):
from tensorflow.python.grappler import tf_optimizer
rewriter_config = rewriter_config_pb2.RewriterConfig()
rewriter_config.optimizers.append("autoparallel")
rewriter_config.auto_parallel.enable = True
rewriter_config.auto_parallel.num_replicas = FLAGS.num_gpus
optimized_graph = tf_optimizer.OptimizeGraph(rewriter_config, metagraph)
metagraph.graph_def.CopyFrom(optimized_graph)
UpdateCollection(metagraph, model)
tutorials/rnn/quickdraw/BUILD deleted 100644 → 0
View file @ 15e0057f
# Description:
# Example classification model on Quick, Draw! dataset.
package(default_visibility = ["//visibility:public"])
licenses(["notice"]) # Apache 2.0
exports_files(["LICENSE"])
py_binary(
name = "train_model",
srcs = [
"train_model.py",
],
srcs_version = "PY2AND3",
deps = [
"//third_party/py/tensorflow",
],
)
py_binary(
name = "create_dataset",
srcs = [
"create_dataset.py",
],
deps = [
"//third_party/py/numpy",
"//third_party/py/tensorflow",
],
)
filegroup(
name = "all_files",
srcs = glob(
["**/*"],
exclude = [
"**/METADATA",
"**/OWNERS",
],
),
visibility = ["//third_party/tensorflow:__subpackages__"],
)
tutorials/rnn/quickdraw/create_dataset.py deleted 100644 → 0
View file @ 15e0057f
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
r"""Creates training and eval data from Quickdraw NDJSON files.
This tool reads the NDJSON files from https://quickdraw.withgoogle.com/data
and converts them into tensorflow.Example stored in TFRecord files.
The tensorflow example will contain 3 features:
shape - contains the shape of the sequence [length, dim] where dim=3.
class_index - the class index of the class for the example.
ink - a length * dim vector of the ink.
It creates disjoint training and evaluation sets.
python create_dataset.py \
--ndjson_path ${HOME}/ndjson \
--output_path ${HOME}/tfrecord
"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import json
import os
import random
import sys
import numpy as np
import tensorflow as tf
def parse_line(ndjson_line):
"""Parse an ndjson line and return ink (as np array) and classname."""
sample = json.loads(ndjson_line)
class_name = sample["word"]
if not class_name:
print ("Empty classname")
return None, None
inkarray = sample["drawing"]
stroke_lengths = [len(stroke[0]) for stroke in inkarray]
total_points = sum(stroke_lengths)
np_ink = np.zeros((total_points, 3), dtype=np.float32)
current_t = 0
if not inkarray:
print("Empty inkarray")
return None, None
for stroke in inkarray:
if len(stroke[0]) != len(stroke[1]):
print("Inconsistent number of x and y coordinates.")
return None, None
for i in [0, 1]:
np_ink[current_t:(current_t + len(stroke[0])), i] = stroke[i]
current_t += len(stroke[0])
np_ink[current_t - 1, 2] = 1 # stroke_end
# Preprocessing.
# 1. Size normalization.
lower = np.min(np_ink[:, 0:2], axis=0)
upper = np.max(np_ink[:, 0:2], axis=0)
scale = upper - lower
scale[scale == 0] = 1
np_ink[:, 0:2] = (np_ink[:, 0:2] - lower) / scale
# 2. Compute deltas.
np_ink[1:, 0:2] -= np_ink[0:-1, 0:2]
np_ink = np_ink[1:, :]
return np_ink, class_name
def convert_data(trainingdata_dir,
observations_per_class,
output_file,
classnames,
output_shards=10,
offset=0):
"""Convert training data from ndjson files into tf.Example in tf.Record.
Args:
trainingdata_dir: path to the directory containin the training data.
The training data is stored in that directory as ndjson files.
observations_per_class: the number of items to load per class.
output_file: path where to write the output.
classnames: array with classnames - is auto created if not passed in.
output_shards: the number of shards to write the output in.
offset: the number of items to skip at the beginning of each file.
Returns:
classnames: the class names as strings. classnames[classes[i]] is the
textual representation of the class of the i-th data point.
"""
def _pick_output_shard():
return random.randint(0, output_shards - 1)
file_handles = []
# Open all input files.
for filename in sorted(tf.gfile.ListDirectory(trainingdata_dir)):
if not filename.endswith(".ndjson"):
print("Skipping", filename)
continue
file_handles.append(
tf.gfile.GFile(os.path.join(trainingdata_dir, filename), "r"))
if offset: # Fast forward all files to skip the offset.
count = 0
for _ in file_handles[-1]:
count += 1
if count == offset:
break
writers = []
for i in range(FLAGS.output_shards):
writers.append(
tf.python_io.TFRecordWriter("%s-%05i-of-%05i" % (output_file, i,
output_shards)))
reading_order = list(range(len(file_handles))) * observations_per_class
random.shuffle(reading_order)
for c in reading_order:
line = file_handles[c].readline()
ink = None
while ink is None:
ink, class_name = parse_line(line)
if ink is None:
print ("Couldn't parse ink from '" + line + "'.")
if class_name not in classnames:
classnames.append(class_name)
features = {}
features["class_index"] = tf.train.Feature(int64_list=tf.train.Int64List(
value=[classnames.index(class_name)]))
features["ink"] = tf.train.Feature(float_list=tf.train.FloatList(
value=ink.flatten()))
features["shape"] = tf.train.Feature(int64_list=tf.train.Int64List(
value=ink.shape))
f = tf.train.Features(feature=features)
example = tf.train.Example(features=f)
writers[_pick_output_shard()].write(example.SerializeToString())
# Close all files
for w in writers:
w.close()
for f in file_handles:
f.close()
# Write the class list.
with tf.gfile.GFile(output_file + ".classes", "w") as f:
for class_name in classnames:
f.write(class_name + "\n")
return classnames
def main(argv):
del argv
classnames = convert_data(
FLAGS.ndjson_path,
FLAGS.train_observations_per_class,
os.path.join(FLAGS.output_path, "training.tfrecord"),
classnames=[],
output_shards=FLAGS.output_shards,
offset=0)
convert_data(
FLAGS.ndjson_path,
FLAGS.eval_observations_per_class,
os.path.join(FLAGS.output_path, "eval.tfrecord"),
classnames=classnames,
output_shards=FLAGS.output_shards,
offset=FLAGS.train_observations_per_class)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.register("type", "bool", lambda v: v.lower() == "true")
parser.add_argument(
"--ndjson_path",
type=str,
default="",
help="Directory where the ndjson files are stored.")
parser.add_argument(
"--output_path",
type=str,
default="",
help="Directory where to store the output TFRecord files.")
parser.add_argument(
"--train_observations_per_class",
type=int,
default=10000,
help="How many items per class to load for training.")
parser.add_argument(
"--eval_observations_per_class",
type=int,
default=1000,
help="How many items per class to load for evaluation.")
parser.add_argument(
"--output_shards",
type=int,
default=10,
help="Number of shards for the output.")
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
tutorials/rnn/quickdraw/train_model.py deleted 100644 → 0
View file @ 15e0057f
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
r"""Binary for training a RNN-based classifier for the Quick, Draw! data.
python train_model.py \
--training_data train_data \
--eval_data eval_data \
--model_dir /tmp/quickdraw_model/ \
--cell_type cudnn_lstm
When running on GPUs using --cell_type cudnn_lstm is much faster.
The expected performance is ~75% in 1.5M steps with the default configuration.
"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import ast
import functools
import sys
import tensorflow as tf
def get_num_classes():
classes = []
with tf.gfile.GFile(FLAGS.classes_file, "r") as f:
classes = [x for x in f]
num_classes = len(classes)
return num_classes
def get_input_fn(mode, tfrecord_pattern, batch_size):
"""Creates an input_fn that stores all the data in memory.
Args:
mode: one of tf.contrib.learn.ModeKeys.{TRAIN, INFER, EVAL}
tfrecord_pattern: path to a TF record file created using create_dataset.py.
batch_size: the batch size to output.
Returns:
A valid input_fn for the model estimator.
"""
def _parse_tfexample_fn(example_proto, mode):
"""Parse a single record which is expected to be a tensorflow.Example."""
feature_to_type = {
"ink": tf.VarLenFeature(dtype=tf.float32),
"shape": tf.FixedLenFeature([2], dtype=tf.int64)
}
if mode != tf.estimator.ModeKeys.PREDICT:
# The labels won't be available at inference time, so don't add them
# to the list of feature_columns to be read.
feature_to_type["class_index"] = tf.FixedLenFeature([1], dtype=tf.int64)
parsed_features = tf.parse_single_example(example_proto, feature_to_type)
labels = None
if mode != tf.estimator.ModeKeys.PREDICT:
labels = parsed_features["class_index"]
parsed_features["ink"] = tf.sparse_tensor_to_dense(parsed_features["ink"])
return parsed_features, labels
def _input_fn():
"""Estimator `input_fn`.
Returns:
A tuple of:
- Dictionary of string feature name to `Tensor`.
- `Tensor` of target labels.
"""
dataset = tf.data.TFRecordDataset.list_files(tfrecord_pattern)
if mode == tf.estimator.ModeKeys.TRAIN:
dataset = dataset.shuffle(buffer_size=10)
dataset = dataset.repeat()
# Preprocesses 10 files concurrently and interleaves records from each file.
dataset = dataset.interleave(
tf.data.TFRecordDataset,
cycle_length=10,
block_length=1)
dataset = dataset.map(
functools.partial(_parse_tfexample_fn, mode=mode),
num_parallel_calls=10)
dataset = dataset.prefetch(10000)
if mode == tf.estimator.ModeKeys.TRAIN:
dataset = dataset.shuffle(buffer_size=1000000)
# Our inputs are variable length, so pad them.
dataset = dataset.padded_batch(
batch_size, padded_shapes=dataset.output_shapes)
features, labels = dataset.make_one_shot_iterator().get_next()
return features, labels
return _input_fn
def model_fn(features, labels, mode, params):
"""Model function for RNN classifier.
This function sets up a neural network which applies convolutional layers (as
configured with params.num_conv and params.conv_len) to the input.
The output of the convolutional layers is given to LSTM layers (as configured
with params.num_layers and params.num_nodes).
The final state of the all LSTM layers are concatenated and fed to a fully
connected layer to obtain the final classification scores.
Args:
features: dictionary with keys: inks, lengths.
labels: one hot encoded classes
mode: one of tf.estimator.ModeKeys.{TRAIN, INFER, EVAL}
params: a parameter dictionary with the following keys: num_layers,
num_nodes, batch_size, num_conv, conv_len, num_classes, learning_rate.
Returns:
ModelFnOps for Estimator API.
"""
def _get_input_tensors(features, labels):
"""Converts the input dict into inks, lengths, and labels tensors."""
# features[ink] is a sparse tensor that is [8, batch_maxlen, 3]
# inks will be a dense tensor of [8, maxlen, 3]
# shapes is [batchsize, 2]
shapes = features["shape"]
# lengths will be [batch_size]
lengths = tf.squeeze(
tf.slice(shapes, begin=[0, 0], size=[params.batch_size, 1]))
inks = tf.reshape(features["ink"], [params.batch_size, -1, 3])
if labels is not None:
labels = tf.squeeze(labels)
return inks, lengths, labels
def _add_conv_layers(inks, lengths):
"""Adds convolution layers."""
convolved = inks
for i in range(len(params.num_conv)):
convolved_input = convolved
if params.batch_norm:
convolved_input = tf.layers.batch_normalization(
convolved_input,
training=(mode == tf.estimator.ModeKeys.TRAIN))
# Add dropout layer if enabled and not first convolution layer.
if i > 0 and params.dropout:
convolved_input = tf.layers.dropout(
convolved_input,
rate=params.dropout,
training=(mode == tf.estimator.ModeKeys.TRAIN))
convolved = tf.layers.conv1d(
convolved_input,
filters=params.num_conv[i],
kernel_size=params.conv_len[i],
activation=None,
strides=1,
padding="same",
name="conv1d_%d" % i)
return convolved, lengths
def _add_regular_rnn_layers(convolved, lengths):
"""Adds RNN layers."""
if params.cell_type == "lstm":
cell = tf.nn.rnn_cell.BasicLSTMCell
elif params.cell_type == "block_lstm":
cell = tf.contrib.rnn.LSTMBlockCell
cells_fw = [cell(params.num_nodes) for _ in range(params.num_layers)]
cells_bw = [cell(params.num_nodes) for _ in range(params.num_layers)]
if params.dropout > 0.0:
cells_fw = [tf.contrib.rnn.DropoutWrapper(cell) for cell in cells_fw]
cells_bw = [tf.contrib.rnn.DropoutWrapper(cell) for cell in cells_bw]
outputs, _, _ = tf.contrib.rnn.stack_bidirectional_dynamic_rnn(
cells_fw=cells_fw,
cells_bw=cells_bw,
inputs=convolved,
sequence_length=lengths,
dtype=tf.float32,
scope="rnn_classification")
return outputs
def _add_cudnn_rnn_layers(convolved):
"""Adds CUDNN LSTM layers."""
# Convolutions output [B, L, Ch], while CudnnLSTM is time-major.
convolved = tf.transpose(convolved, [1, 0, 2])
lstm = tf.contrib.cudnn_rnn.CudnnLSTM(
num_layers=params.num_layers,
num_units=params.num_nodes,
dropout=params.dropout if mode == tf.estimator.ModeKeys.TRAIN else 0.0,
direction="bidirectional")
outputs, _ = lstm(convolved)
# Convert back from time-major outputs to batch-major outputs.
outputs = tf.transpose(outputs, [1, 0, 2])
return outputs
def _add_rnn_layers(convolved, lengths):
"""Adds recurrent neural network layers depending on the cell type."""
if params.cell_type != "cudnn_lstm":
outputs = _add_regular_rnn_layers(convolved, lengths)
else:
outputs = _add_cudnn_rnn_layers(convolved)
# outputs is [batch_size, L, N] where L is the maximal sequence length and N
# the number of nodes in the last layer.
mask = tf.tile(
tf.expand_dims(tf.sequence_mask(lengths, tf.shape(outputs)[1]), 2),
[1, 1, tf.shape(outputs)[2]])
zero_outside = tf.where(mask, outputs, tf.zeros_like(outputs))
outputs = tf.reduce_sum(zero_outside, axis=1)
return outputs
def _add_fc_layers(final_state):
"""Adds a fully connected layer."""
return tf.layers.dense(final_state, params.num_classes)
# Build the model.
inks, lengths, labels = _get_input_tensors(features, labels)
convolved, lengths = _add_conv_layers(inks, lengths)
final_state = _add_rnn_layers(convolved, lengths)
logits = _add_fc_layers(final_state)
# Add the loss.
cross_entropy = tf.reduce_mean(
tf.nn.sparse_softmax_cross_entropy_with_logits(
labels=labels, logits=logits))
# Add the optimizer.
train_op = tf.contrib.layers.optimize_loss(
loss=cross_entropy,
global_step=tf.train.get_global_step(),
learning_rate=params.learning_rate,
optimizer="Adam",
# some gradient clipping stabilizes training in the beginning.
clip_gradients=params.gradient_clipping_norm,
summaries=["learning_rate", "loss", "gradients", "gradient_norm"])
# Compute current predictions.
predictions = tf.argmax(logits, axis=1)
return tf.estimator.EstimatorSpec(
mode=mode,
predictions={"logits": logits, "predictions": predictions},
loss=cross_entropy,
train_op=train_op,
eval_metric_ops={"accuracy": tf.metrics.accuracy(labels, predictions)})
def create_estimator_and_specs(run_config):
"""Creates an Experiment configuration based on the estimator and input fn."""
model_params = tf.contrib.training.HParams(
num_layers=FLAGS.num_layers,
num_nodes=FLAGS.num_nodes,
batch_size=FLAGS.batch_size,
num_conv=ast.literal_eval(FLAGS.num_conv),
conv_len=ast.literal_eval(FLAGS.conv_len),
num_classes=get_num_classes(),
learning_rate=FLAGS.learning_rate,
gradient_clipping_norm=FLAGS.gradient_clipping_norm,
cell_type=FLAGS.cell_type,
batch_norm=FLAGS.batch_norm,
dropout=FLAGS.dropout)
estimator = tf.estimator.Estimator(
model_fn=model_fn,
config=run_config,
params=model_params)
train_spec = tf.estimator.TrainSpec(input_fn=get_input_fn(
mode=tf.estimator.ModeKeys.TRAIN,
tfrecord_pattern=FLAGS.training_data,
batch_size=FLAGS.batch_size), max_steps=FLAGS.steps)
eval_spec = tf.estimator.EvalSpec(input_fn=get_input_fn(
mode=tf.estimator.ModeKeys.EVAL,
tfrecord_pattern=FLAGS.eval_data,
batch_size=FLAGS.batch_size))
return estimator, train_spec, eval_spec
def main(unused_args):
estimator, train_spec, eval_spec = create_estimator_and_specs(
run_config=tf.estimator.RunConfig(
model_dir=FLAGS.model_dir,
save_checkpoints_secs=300,
save_summary_steps=100))
tf.estimator.train_and_evaluate(estimator, train_spec, eval_spec)
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.register("type", "bool", lambda v: v.lower() == "true")
parser.add_argument(
"--training_data",
type=str,
default="",
help="Path to training data (tf.Example in TFRecord format)")
parser.add_argument(
"--eval_data",
type=str,
default="",
help="Path to evaluation data (tf.Example in TFRecord format)")
parser.add_argument(
"--classes_file",
type=str,
default="",
help="Path to a file with the classes - one class per line")
parser.add_argument(
"--num_layers",
type=int,
default=3,
help="Number of recurrent neural network layers.")
parser.add_argument(
"--num_nodes",
type=int,
default=128,
help="Number of node per recurrent network layer.")
parser.add_argument(
"--num_conv",
type=str,
default="[48, 64, 96]",
help="Number of conv layers along with number of filters per layer.")
parser.add_argument(
"--conv_len",
type=str,
default="[5, 5, 3]",
help="Length of the convolution filters.")
parser.add_argument(
"--cell_type",
type=str,
default="lstm",
help="Cell type used for rnn layers: cudnn_lstm, lstm or block_lstm.")
parser.add_argument(
"--batch_norm",
type="bool",
default="False",
help="Whether to enable batch normalization or not.")
parser.add_argument(
"--learning_rate",
type=float,
default=0.0001,
help="Learning rate used for training.")
parser.add_argument(
"--gradient_clipping_norm",
type=float,
default=9.0,
help="Gradient clipping norm used during training.")
parser.add_argument(
"--dropout",
type=float,
default=0.3,
help="Dropout used for convolutions and bidi lstm layers.")
parser.add_argument(
"--steps",
type=int,
default=100000,
help="Number of training steps.")
parser.add_argument(
"--batch_size",
type=int,
default=8,
help="Batch size to use for training/evaluation.")
parser.add_argument(
"--model_dir",
type=str,
default="",
help="Path for storing the model checkpoints.")
parser.add_argument(
"--self_test",
type="bool",
default="False",
help="Whether to enable batch normalization or not.")
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment