Skip to content

Instantly share code, notes, and snippets.

@jasonsalas
Created November 20, 2019 03:40
Show Gist options
  • Save jasonsalas/793a451a2c0a4f7fa6766bc29c4c0544 to your computer and use it in GitHub Desktop.
Save jasonsalas/793a451a2c0a4f7fa6766bc29c4c0544 to your computer and use it in GitHub Desktop.
Generate Depeche Mode lyrics automatically with an LSTM neural network
import numpy as np
import sys
from keras.models import Sequential
from keras.layers import Dense, Dropout, LSTM
from keras.callbacks import ModelCheckpoint
from keras.utils import to_categorical
filename = 'dm_lyrics.txt'
raw_text = open(filename, 'r', encoding='utf-8').read()
raw_text = raw_text.lower()
# character-to-integer mapping
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 vocabulary: ', n_vocab)
# prepare the dataset of input to output pairs encoded as integers
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)
'''
sequence input format expected by the LSTM-based neural network:
[samples, timesteps, features]
'''
# reshape X to the expected input format
X = np.reshape(dataX, (n_patterns, seq_length, 1))
# normalization
X = X / float(n_vocab)
# one-hot encoding of the output
y = to_categorical(dataY)
print(X.shape, y.shape)
# define the LSTM model
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')
model.summary()
filepath = 'weights_improvement_{epoch:02d}_{loss:.4f}.h5'
checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, save_best_only=True, mode='min')
model.fit(X, y, epochs=3, batch_size=128, callbacks=[checkpoint])
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment