Skip to content

Fixed the random number generators. #83

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
nouiz merged 2 commits into from
May 11, 2015
Merged

Fixed the random number generators. #83

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
423f54d
fixed the random number generators.
caglar May 11, 2015
29b167a
converted the code to use numpy random number generator.
caglar May 11, 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
14 changes: 8 additions & 6 deletions code/lstm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,6 @@
'''
from collections import OrderedDict
import cPickle as pkl
import random
import sys
import time

Expand All @@ -17,6 +16,9 @@

datasets = {'imdb': (imdb.load_data, imdb.prepare_data)}

# Set the random number generators' seeds for consistency
SEED = 123
numpy.random.seed(SEED)

def numpy_floatX(data):
return numpy.asarray(data, dtype=config.floatX)
Expand All @@ -30,7 +32,7 @@ def get_minibatches_idx(n, minibatch_size, shuffle=False):
idx_list = numpy.arange(n, dtype="int32")

if shuffle:
random.shuffle(idx_list)
numpy.random.shuffle(idx_list)

minibatches = []
minibatch_start = 0
Expand Down Expand Up @@ -303,7 +305,7 @@ def rmsprop(lr, tparams, grads, x, mask, y, cost):


def build_model(tparams, options):
trng = RandomStreams(1234)
trng = RandomStreams(SEED)

# Used for dropout.
use_noise = theano.shared(numpy_floatX(0.))
Expand Down Expand Up @@ -401,7 +403,7 @@ def train_lstm(
noise_std=0.,
use_dropout=True, # if False slightly faster, but worst test error
# This frequently need a bigger model.
reload_model="", # Path to a saved model we want to start from.
reload_model=None, # Path to a saved model we want to start from.
test_size=-1, # If >0, we keep only this number of test example.
):

Expand All @@ -419,7 +421,7 @@ def train_lstm(
# size example. So we must select a random selection of the
# examples.
idx = numpy.arange(len(test[0]))
random.shuffle(idx)
numpy.random.shuffle(idx)
idx = idx[:test_size]
test = ([test[0][n] for n in idx], [test[1][n] for n in idx])

Expand Down Expand Up @@ -468,6 +470,7 @@ def train_lstm(
print "%d train examples" % len(train[0])
print "%d valid examples" % len(valid[0])
print "%d test examples" % len(test[0])

history_errs = []
best_p = None
bad_count = 0
Expand Down Expand Up @@ -585,7 +588,6 @@ def train_lstm(
if __name__ == '__main__':
# See function train for all possible parameter and there definition.
train_lstm(
#reload_model="lstm_model.npz",
max_epochs=100,
test_size=500,
)