Skip to content

Update/speed up to the lstm code #71

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 7 commits into from
Feb 3, 2015
+84 −52
Merged

Update/speed up to the lstm code #71

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
9018392
pep8
nouiz Jan 14, 2015
906a30e
Small comment update
nouiz Jan 16, 2015
26914e4
Add a parameter to make the test set smaller
nouiz Jan 16, 2015
ec112fa
use floatX for lstm
nouiz Jan 21, 2015
6a32f03
sort dataset by length to speed up error computation
nouiz Jan 31, 2015
2f14fbf
Try to repair travis failure to install our dependency
nouiz Jan 31, 2015
6e38783
Keep random subset of the test set
nouiz Feb 3, 2015
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
pep8
  • Loading branch information
@nouiz
nouiz committed Jan 31, 2015
commit 9018392ea487e37f6620cef6222bc3870f2a3dfb
2 changes: 0 additions & 2 deletions code/imdb.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,8 +4,6 @@

import numpy

import theano


def prepare_data(seqs, labels, maxlen=None):
"""Create the matrices from the datasets.
Expand Down
24 changes: 12 additions & 12 deletions code/lstm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,6 @@
Build a tweet sentiment analyzer
'''
from collections import OrderedDict
import copy
import cPickle as pkl
import random
import sys
Expand Down Expand Up @@ -159,8 +158,8 @@ def lstm_layer(tparams, state_below, options, prefix='lstm', mask=None):

def _slice(_x, n, dim):
if _x.ndim == 3:
return _x[:, :, n*dim:(n+1)*dim]
return _x[:, n*dim:(n+1)*dim]
return _x[:, :, n * dim:(n + 1) * dim]
return _x[:, n * dim:(n + 1) * dim]

def _step(m_, x_, h_, c_):
preact = tensor.dot(h_, tparams[_p(prefix, 'U')])
Expand Down Expand Up @@ -243,7 +242,7 @@ def adadelta(lr, tparams, grads, x, mask, y, cost):
rg2up = [(rg2, 0.95 * rg2 + 0.05 * (g ** 2))
for rg2, g in zip(running_grads2, grads)]

f_grad_shared = theano.function([x, mask, y], cost, updates=zgup+rg2up,
f_grad_shared = theano.function([x, mask, y], cost, updates=zgup + rg2up,
name='adadelta_f_grad_shared')

updir = [-tensor.sqrt(ru2 + 1e-6) / tensor.sqrt(rg2 + 1e-6) * zg
Expand All @@ -254,7 +253,7 @@ def adadelta(lr, tparams, grads, x, mask, y, cost):
for ru2, ud in zip(running_up2, updir)]
param_up = [(p, p + ud) for p, ud in zip(tparams.values(), updir)]

f_update = theano.function([lr], [], updates=ru2up+param_up,
f_update = theano.function([lr], [], updates=ru2up + param_up,
on_unused_input='ignore',
name='adadelta_f_update')

Expand Down Expand Up @@ -289,7 +288,7 @@ def rmsprop(lr, tparams, grads, x, mask, y, cost):
running_grads2)]
param_up = [(p, p + udn[1])
for p, udn in zip(tparams.values(), updir_new)]
f_update = theano.function([lr], [], updates=updir_new+param_up,
f_update = theano.function([lr], [], updates=updir_new + param_up,
on_unused_input='ignore',
name='rmsprop_f_update')

Expand Down Expand Up @@ -321,7 +320,7 @@ def build_model(tparams, options):
if options['use_dropout']:
proj = dropout_layer(proj, use_noise, trng)

pred = tensor.nnet.softmax(tensor.dot(proj, tparams['U'])+tparams['b'])
pred = tensor.nnet.softmax(tensor.dot(proj, tparams['U']) + tparams['b'])

f_pred_prob = theano.function([x, mask], pred, name='f_pred_prob')
f_pred = theano.function([x, mask], pred.argmax(axis=1), name='f_pred')
Expand Down Expand Up @@ -408,7 +407,7 @@ def train_lstm(
train, valid, test = load_data(n_words=n_words, valid_portion=0.05,
maxlen=maxlen)

ydim = numpy.max(train[1])+1
ydim = numpy.max(train[1]) + 1

model_options['ydim'] = ydim

Expand All @@ -432,7 +431,7 @@ def train_lstm(
if decay_c > 0.:
decay_c = theano.shared(numpy.float32(decay_c), name='decay_c')
weight_decay = 0.
weight_decay += (tparams['U']**2).sum()
weight_decay += (tparams['U'] ** 2).sum()
weight_decay *= decay_c
cost += weight_decay

Expand Down Expand Up @@ -460,9 +459,9 @@ def train_lstm(
bad_count = 0

if validFreq == -1:
validFreq = len(train[0])/batch_size
validFreq = len(train[0]) / batch_size
if saveFreq == -1:
saveFreq = len(train[0])/batch_size
saveFreq = len(train[0]) / batch_size

uidx = 0 # the number of update done
estop = False # early stop
Expand Down Expand Up @@ -514,7 +513,8 @@ def train_lstm(
if numpy.mod(uidx, validFreq) == 0:
use_noise.set_value(0.)
train_err = pred_error(f_pred, prepare_data, train, kf)
valid_err = pred_error(f_pred, prepare_data, valid, kf_valid)
valid_err = pred_error(f_pred, prepare_data, valid,
kf_valid)
test_err = pred_error(f_pred, prepare_data, test, kf_test)

history_errs.append([valid_err, test_err])
Expand Down