Created
January 1, 2019 10:43
-
-
Save arshjat/24f37f9354a99f5f0bef1142192f0f36 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def build_model1(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32): | |
file_path = "best_model.hdf5" | |
check_point = ModelCheckpoint(file_path, monitor = "val_loss", verbose = 1, | |
save_best_only = True, mode = "min") | |
early_stop = EarlyStopping(monitor = "val_loss", mode = "min", patience = 3) | |
inp = Input(shape = (max_len,)) | |
x = Embedding(30001, embed_size, weights = [embedding_matrix], trainable = False)(inp) | |
x1 = SpatialDropout1D(spatial_dr)(x) | |
x_gru = Bidirectional(CuDNNGRU(units, return_sequences = True))(x1) | |
x1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_gru) | |
avg_pool1_gru = GlobalAveragePooling1D()(x1) | |
max_pool1_gru = GlobalMaxPooling1D()(x1) | |
x3 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_gru) | |
avg_pool3_gru = GlobalAveragePooling1D()(x3) | |
max_pool3_gru = GlobalMaxPooling1D()(x3) | |
x_lstm = Bidirectional(CuDNNLSTM(units, return_sequences = True))(x1) | |
x1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_lstm) | |
avg_pool1_lstm = GlobalAveragePooling1D()(x1) | |
max_pool1_lstm = GlobalMaxPooling1D()(x1) | |
x3 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_lstm) | |
avg_pool3_lstm = GlobalAveragePooling1D()(x3) | |
max_pool3_lstm = GlobalMaxPooling1D()(x3) | |
x = concatenate([avg_pool1_gru, max_pool1_gru, avg_pool3_gru, max_pool3_gru, | |
avg_pool1_lstm, max_pool1_lstm, avg_pool3_lstm, max_pool3_lstm]) | |
x = BatchNormalization()(x) | |
x = Dropout(dr)(Dense(dense_units, activation='relu') (x)) | |
x = BatchNormalization()(x) | |
x = Dropout(dr)(Dense(int(dense_units / 2), activation='relu') (x)) | |
x = Dense(1, activation = "sigmoid")(x) | |
model = Model(inputs = inp, outputs = x) | |
model.compile(loss = "binary_crossentropy", optimizer = Adam(lr = lr, decay = lr_d), metrics = ["accuracy"]) | |
history = model.fit(X_train, y, batch_size = 128, epochs = 10, validation_split=0.1, | |
verbose = 1, callbacks = [check_point, early_stop]) | |
model = load_model(file_path) | |
return model |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
can you please write about the defination of embedding_matrix(Line 8)