main-lm.py

"""
    Tests NPRN for Language Modeling
    Borrowed from https://github.com/pytorch/examples/tree/master/word_language_model
"""

import argparse
import time
import math
import os
import torch
from torch.autograd import Variable
import torch.nn as nn

from nprn import GaussianNPRNLanguageModel

import data

parser = argparse.ArgumentParser(description='PyTorch Wikitext-2 Gaussian NPRN Language Model')
parser.add_argument('--data', type=str, default='./data/wikitext-2',
                    help='location of the data corpus')
parser.add_argument('--model', type=str, default='LSTM',
                    help='type of recurrent net (RNN_TANH, RNN_RELU, LSTM, GRU)')
parser.add_argument('--emsize', type=int, default=200,
                    help='size of word embeddings')
parser.add_argument('--nhid', type=int, default=200,
                    help='number of hidden units per layer')
parser.add_argument('--lr', type=float, default=20,
                    help='initial learning rate')
parser.add_argument('--clip', type=float, default=0.25,
                    help='gradient clipping')
parser.add_argument('--epochs', type=int, default=40,
                    help='upper epoch limit')
parser.add_argument('--batch_size', type=int, default=20, metavar='N',
                    help='batch size')
parser.add_argument('--bptt', type=int, default=35,
                    help='sequence length')
parser.add_argument('--dropout', type=float, default=0.2,
                    help='dropout applied to layers (0 = no dropout)')

# this will only be valid with NPRN (can't have tied weights)
parser.add_argument('--tied', action='store_true',
                    help='tie the word embedding and softmax weights')
parser.add_argument('--seed', type=int, default=1111,
                    help='random seed')
parser.add_argument('--cuda', action='store_true',
                    help='use CUDA')
parser.add_argument('--log-interval', type=int, default=200, metavar='N',
                    help='report interval')
parser.add_argument('--save', type=str, default='model.pt',
                    help='path to save the final model')
args = parser.parse_args()

# Set the random seed manually for reproducibility.
torch.manual_seed(args.seed)
if torch.cuda.is_available():
    args.cuda = True


###############################################################################
# Load data
###############################################################################

corpus = data.Corpus(args.data)

# Starting from sequential data, batchify arranges the dataset into columns.
# For instance, with the alphabet as the sequence and batch size 4, we'd get
# ┌ a g m s ┐
# │ b h n t │
# │ c i o u │
# │ d j p v │
# │ e k q w │
# └ f l r x ┘.
# These columns are treated as independent by the model, which means that the
# dependence of e. g. 'g' on 'f' can not be learned, but allows more efficient
# batch processing.

def batchify(data, bsz):
    # Work out how cleanly we can divide the dataset into bsz parts.
    nbatch = data.size(0) // bsz
    # Trim off any extra elements that wouldn't cleanly fit (remainders).
    data = data.narrow(0, 0, nbatch * bsz)
    # Evenly divide the data across the bsz batches.
    data = data.view(bsz, -1).t().contiguous()
    return data

eval_batch_size = 10
train_data = batchify(corpus.train, args.batch_size)
val_data = batchify(corpus.valid, eval_batch_size)
test_data = batchify(corpus.test, eval_batch_size)

###############################################################################
# Build the model
###############################################################################

ntokens = len(corpus.dictionary)
model = GaussianNPRNLanguageModel(ntokens, args.emsize, args.nhid)

if args.cuda:
    model = model.cuda()

criterion = nn.BCELoss(size_average=False)
nllcriterion = nn.CrossEntropyLoss()
###############################################################################
# Training code
###############################################################################

def repackage_hidden(h):
    """Wraps hidden states in new Tensors, to detach them from their history."""
    if isinstance(h, Variable):
        return h.detach()
    elif isinstance(h,tuple):
        return tuple(repackage_hidden(v) for v in h)
    else:
        raise ValueError("hidden can't be %s" % type(h))


# get_batch subdivides the source data into chunks of length args.bptt.
# If source is equal to the example output of the batchify function, with
# a bptt-limit of 2, we'd get the following two Variables for i = 0:
# ┌ a g m s ┐ ┌ b h n t ┐
# └ b h n t ┘ └ c i o u ┘
# Note that despite the name of the function, the subdivison of data is not
# done along the batch dimension (i.e. dimension 1), since that was handled
# by the batchify function. The chunks are along dimension 0, corresponding
# to the seq_len dimension in the LSTM.

def get_batch(source, i, ntok):
    seq_len = min(args.bptt, len(source) - 1 - i)
    data = source[i:i+seq_len]
    if args.cuda:
        data = data.cuda()
    target = source[i+1:i+1+seq_len].view(-1)
    target_onehot = torch.FloatTensor(target.size(0), ntok)
    target_onehot.zero_()
    target_onehot.scatter_(1, target.unsqueeze(1), 1)
    if args.cuda:
        target_onehot = target_onehot.cuda()
        target = target.cuda()
    target_onehot = Variable(target_onehot)
    target = Variable(target)
    return data, target, target_onehot


def evaluate(data_source):
    model.eval()
    total_loss = 0.
    ntokens = len(corpus.dictionary)
    hidden = model.init_hidden(eval_batch_size)
    with torch.no_grad():
        for i in range(0, data_source.size(0) - 1, args.bptt):
            data, targets, _ = get_batch(data_source, i, ntokens)
            output, hidden = model.logits(data, hidden)
            output_flat = output[0].view(-1, ntokens)
            total_loss += nllcriterion(output_flat, targets).item()
            hidden = repackage_hidden(hidden)
    return total_loss / len(data_source)


def train(optimizer):
    # Turn on training mode which enables dropout.
    model.train()
    total_loss = 0.
    total_nll_loss = 0.
    start_time = time.time()
    ntokens = len(corpus.dictionary)
    hidden = model.init_hidden(args.batch_size)
    for batch, i in enumerate(range(0, train_data.size(0) - 1, args.bptt)):
        data, targets, targets_onehot = get_batch(train_data, i, ntokens)
        # Starting each batch, we detach the hidden state from how it was previously produced.
        # If we didn't, the model would try backpropagating all the way to start of the dataset.
        data = Variable(data)
        hidden = repackage_hidden(hidden)
        optimizer.zero_grad()
        output, hidden = model(data, hidden)
        output_m, output_s = output
        output_flat = output_m.view(-1, ntokens)
        loss = criterion(output_flat, targets_onehot)
        loss.backward()
        nll_loss = nllcriterion(output_flat, targets)

        optimizer.step()

        total_loss += loss.data.item()
        total_nll_loss += nll_loss.data.item()


        if batch % args.log_interval == 0 and batch > 0:
            cur_loss = total_loss / args.log_interval
            cur_nll_loss = total_nll_loss / args.log_interval
            elapsed = time.time() - start_time
            print('| epoch {:3d} | {:5d}/{:5d} batches | lr {:02.2f} | ms/batch {:5.2f} | '
                    'loss {:5.2f} | ppl {:8.2f}'.format(
                epoch, batch, len(train_data) // args.bptt, lr,
                elapsed * 1000 / args.log_interval, cur_loss, math.exp(cur_nll_loss) ))
            total_loss = 0
            total_nll_loss = 0
            start_time = time.time()


# Loop over epochs.
lr = args.lr
best_val_loss = None

# At any point you can hit Ctrl + C to break out of training early.
try:
    optimizer = torch.optim.Adagrad(model.parameters())
    for epoch in range(1, args.epochs+1):
        epoch_start_time = time.time()
        train(optimizer)
        val_loss = evaluate(val_data)
        print('-' * 89)
        print('| end of epoch {:3d} | time: {:5.2f}s | valid loss {:5.2f} | '
                'valid ppl {:8.2f}'.format(epoch, (time.time() - epoch_start_time),
                                           val_loss, math.exp(val_loss)))
        print('-' * 89)
        # Save the model if the validation loss is the best we've seen so far.
        if not best_val_loss or val_loss < best_val_loss:
            with open(args.save, 'wb') as f:
                torch.save(model, f)
            best_val_loss = val_loss
        else:
            # Anneal the learning rate if no improvement has been seen in the validation dataset.
            lr /= 4.0
except KeyboardInterrupt:
    print('-' * 89)
    print('Exiting from training early')

# Load the best saved model.
with open(args.save, 'rb') as f:
    model = torch.load(f)
    # after load the rnn params are not a continuous chunk of memory
    # this makes them a continuous chunk, and will speed up forward pass
    model.rnn.flatten_parameters()

# Run on test data.
test_loss = evaluate(test_data)
print('=' * 89)
print('| End of training | test loss {:5.2f} | test ppl {:8.2f}'.format(
    test_loss, math.exp(test_loss)))
print('=' * 89)