neon IMDB sentiment classification implementation

readme.md

##Model

This is an implementation of a LSTM model to solve the IMDB sentiment classification task.

Model script

The model run script is included below (imdb_lstm.py).

Trained weights

The trained weights file can be downloaded from AWS using the following link: imdb_lstm.p.

Performance

This model is acheiving 94.3% and 84.5% accuracy on the training and validation set, respectively.

Instructions

Neon version: commit SHA e7ab2c2e2. Make sure that your local repo is synced to this commit and run the installation procedure before proceeding.

If neon is installed into a virtualenv, make sure that it is activated before running the commands below. Also, the commands below use the GPU backend by default so add -b cpu if you are running on a system without a compatible GPU.

To train the model from scratch for 2 epochs, use the command:

python imdb_lstm.py -b gpu -e 2 -eval 1 -r 0 -s imdb_lstm.p

Additional options are available to add features like saving checkpoints and displaying logging information, use the --help option for details.

Citation

When Are Tree Structures Necessary for Deep Learning of Representations?
Jiwei Li, Dan Jurafsky and Eduard Hovy
http://arxiv.org/pdf/1503.00185v1.pdf

imdb_lstm.py

#!/usr/bin/env python
# ----------------------------------------------------------------------------
# Copyright 2015 Nervana Systems Inc.
# 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.
# ----------------------------------------------------------------------------
"""
Example that trains an LSTM or GRU networks for sentiment analysis
Reference:
   See J.Li et al, EMNLP2015 - http://arxiv.org/pdf/1503.00185v5.pdf
$ python examples/imdb_lstm.py -e 2 -eval 1 --rlayer_type lstm
"""

from neon.backends import gen_backend
from neon.data.dataloaders import load_imdb
from neon.data.dataiterator import ArrayIterator
from neon.data.text_preprocessing import pad_data
from neon.initializers import Uniform, GlorotUniform
from neon.layers import (GeneralizedCost, LSTM, Affine, Dropout, LookupTable,
                         RecurrentSum, Recurrent, DeepBiLSTM, DeepBiRNN)
from neon.models import Model
from neon.optimizers import Adagrad
from neon.transforms import Logistic, Tanh, Softmax, CrossEntropyMulti, Accuracy
from neon.callbacks.callbacks import Callbacks
from neon.util.argparser import NeonArgparser, extract_valid_args

# parse the command line arguments
parser = NeonArgparser(__doc__)
parser.add_argument('--rlayer_type', default='lstm',
                    choices=['bilstm', 'lstm', 'birnn', 'rnn'],
                    help='type of recurrent layer to use (lstm, bilstm, rnn, birnn)')

args = parser.parse_args(gen_be=False)

# hyperparameters from the reference
args.batch_size = 128 
gradient_clip_value = 15
vocab_size = 20000
sentence_length = 128 
embedding_dim = 128 
hidden_size = 128 
reset_cells = True

# setup backend
be = gen_backend(**extract_valid_args(args, gen_backend))

# make dataset
path = load_imdb(path=args.data_dir)
(X_train, y_train), (X_test, y_test), nclass = pad_data(path,
                                                        vocab_size=vocab_size,
                                                        sentence_length=sentence_length)

print "Vocab size - ", vocab_size
print "Sentence Length - ", sentence_length
print "# of train sentences", X_train.shape[0]
print "# of test sentence", X_test.shape[0]

train_set = ArrayIterator(X_train, y_train, nclass=2)
valid_set = ArrayIterator(X_test, y_test, nclass=2)
# weight initialization
uni = Uniform(low=-0.1/embedding_dim, high=0.1/embedding_dim)
g_uni = GlorotUniform()

if args.rlayer_type == 'lstm':
    rlayer = LSTM(hidden_size, g_uni, activation=Tanh(),
                  gate_activation=Logistic(), reset_cells=True)
elif args.rlayer_type == 'bilstm':
    rlayer = DeepBiLSTM(hidden_size, g_uni, activation=Tanh(), depth=1,
                        gate_activation=Logistic(), reset_cells=True)
elif args.rlayer_type == 'rnn':
    rlayer = Recurrent(hidden_size, g_uni, activation=Tanh(), reset_cells=True)
elif args.rlayer_type == 'birnn':
    rlayer = DeepBiRNN(hidden_size, g_uni, activation=Tanh(), depth=1, reset_cells=True)


layers = [
    LookupTable(vocab_size=vocab_size, embedding_dim=embedding_dim, init=uni),
    rlayer,
    RecurrentSum(),
    Dropout(keep=0.5),
    Affine(2, g_uni, bias=g_uni, activation=Softmax())
]

model = Model(layers=layers)

cost = GeneralizedCost(costfunc=CrossEntropyMulti(usebits=True))
optimizer = Adagrad(learning_rate=0.01, gradient_clip_value=gradient_clip_value)

# configure callbacks
callbacks = Callbacks(model, eval_set=valid_set, **args.callback_args)

# train model
model.fit(train_set, optimizer=optimizer, num_epochs=args.epochs, cost=cost, callbacks=callbacks)

# eval model
print "Train Accuracy - ", 100 * model.eval(train_set, metric=Accuracy())
print "Test  Accuracy - ", 100 * model.eval(valid_set, metric=Accuracy())