I recently followed a tutorial to make a RNN to generate text: I copied the python code exactly and sort of understand it as well. My model has been trained for 20 epochs and it produces a long repetitive loop of 3 words:
"and the wour and the wour and the wour..."
I have read in Andrej Kaparthy's blog that changing the temperature of the RNN changes its confidence:
Decreasing the temperature from 1 to some lower number (e.g. 0.5) makes the RNN more confident, but also more conservative in its samples.
I want to change this temperature level to reduce the confidence of the RNN in order for it to create new patterns, but as this is my first Machine Learning project I don't know how.
Here is my Python/keras code:
Generating Text File:
# Generate Text
import sys
import numpy
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout
from keras.layers import LSTM
from keras.callbacks import ModelCheckpoint
from keras.utils import np_utils
filename = "king_lear.txt"
raw_text = open(filename).read()
raw_text = raw_text.lower()
chars = sorted(list(set(raw_text)))
char_to_int = dict((c, i) for i, c in enumerate(chars))
int_to_char = dict((i, c) for i, c in enumerate(chars))
n_chars = len(raw_text)
n_vocab = len(chars)
print "Total Characters: ", n_chars
print "Total Vocab: ", n_vocab
seq_length = 100
dataX = []
dataY = []
for i in range(0, n_chars - seq_length, 1):
seq_in = raw_text[i:i + seq_length]
seq_out = raw_text[i + seq_length]
dataX.append([char_to_int[char] for char in seq_in])
dataY.append(char_to_int[seq_out])
n_patterns = len(dataX)
print "Total Patterns: ", n_patterns
X = numpy.reshape(dataX, (n_patterns, seq_length, 1))
X = X / float(n_vocab)
y = np_utils.to_categorical(dataY)
model = Sequential()
model.add(LSTM(256, input_shape=(X.shape[1], X.shape[2]), return_sequences=True))
model.add(Dropout(0.2))
model.add(LSTM(256))
model.add(Dropout(0.2))
model.add(Dense(y.shape[1], activation='softmax'))
filename = "weights-improvement-08-2.0298-bigger.hdf5"
model.load_weights(filename)
model.compile(loss='categorical_crossentropy', optimizer='adam')
start = numpy.random.randint(0, len(dataX)-1)
pattern = dataX[start]
print "Seed:"
print "\"", ''.join([int_to_char[value] for value in pattern]), "\""
for i in range(60):
x = numpy.reshape(pattern, (1, len(pattern), 1))
x = x / float(n_vocab)
prediction = model.predict(x, verbose=0)
index = numpy.argmax(prediction)
result = int_to_char[index]
seq_in = [int_to_char[value] for value in pattern]
sys.stdout.write(result)
pattern.append(index)
pattern = pattern[1:len(pattern)]
print "\nDone."
Learning File:
# Learn Sentences
import numpy
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout
from keras.layers import LSTM
from keras.callbacks import ModelCheckpoint
from keras.utils import np_utils
filename = "king_lear.txt"
raw_text = open(filename).read()
raw_text = raw_text.lower()
chars = sorted(list(set(raw_text)))
char_to_int = dict((c, i) for i, c in enumerate(chars))
n_chars = len(raw_text)
n_vocab = len(chars)
print "Total Characters: ", n_chars
print "Total Vocab: ", n_vocab
seq_length = 100
dataX = []
dataY = []
for i in range(0, n_chars - seq_length, 1):
seq_in = raw_text[i:i + seq_length]
seq_out = raw_text[i + seq_length]
dataX.append([char_to_int[char] for char in seq_in])
dataY.append(char_to_int[seq_out])
n_patterns = len(dataX)
print "Total Patterns: ", n_patterns
X = numpy.reshape(dataX, (n_patterns, seq_length, 1))
X = X / float(n_vocab)
y = np_utils.to_categorical(dataY)
model = Sequential()
model.add(LSTM(256, input_shape=(X.shape[1], X.shape[2]), return_sequences=True))
model.add(Dropout(0.2))
model.add(LSTM(256))
model.add(Dropout(0.2))
model.add(Dense(y.shape[1], activation='softmax'))
model.compile(loss='categorical_crossentropy', optimizer='adam')
filepath="weights-improvement-{epoch:02d}-{loss:.4f}-bigger.hdf5"
checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, save_best_only=True, mode='min')
callbacks_list = [checkpoint]
model.fit(X, y, epochs=50, batch_size=64, callbacks=callbacks_list)
Please help me to do this. If there is anything wrong with this post, don't hesitate to correct me as it is my first question. Thank you very much.