implemented build_model function

backend/IT18218640/train.py

@@ -5,3 +5,47 @@ from sklearn.model_selection import train_test_split
 
 DATA_PATH = "data.json"
 SAVE_MODEL_PATH = "model.h5"
+
+LEARNING_RATE = 0.0001
+EPOCHS = 40
+BATCH_SIZE = 32
+
+NUM_KEYWORDS = 10
+
+
+def build_model(input_shape, learning_rate, error="sparse_categorical_crossentropy"):
+
+    # build network
+    model = keras.Sequential()
+
+    # copy layer 1
+    model.add(keras.layers.Conv2D(64, (3, 3), activation="relu", input_shape=input_shape, kernel_regularizer=keras.regularizers.l2(0.001)))
+    model.add(keras.layers.BatchNormalization())
+    model.add(keras.layers.MaxPool2D((3, 3), strides=(2, 2), padding="same"))
+
+    # copy layer 2
+    model.add(keras.layers.Conv2D(32, (3, 3), activation="relu", kernel_regularizer=keras.regularizers.l2(0.001)))
+    model.add(keras.layers.BatchNormalization())
+    model.add(keras.layers.MaxPool2D((3, 3), strides=(2, 2), padding="same"))
+
+    # copy layer 3
+    model.add(keras.layers.Conv2D(32, (2, 2), activation="relu", kernel_regularizer=keras.regularizers.l2(0.001)))
+    model.add(keras.layers.BatchNormalization())
+    model.add(keras.layers.MaxPool2D((2, 2), strides=(2, 2), padding="same"))
+
+    # flatten the output feed it into a dense layer
+    model.add(keras.layers.Flatten())
+    model.add(keras.layers.Dense(64, activation="relu"))
+    model.add(keras.layers.Dropout(0.3))
+
+    # softmax classifier
+    model.add(keras.layers.Dense(NUM_KEYWORDS, activation="softmax"))
+
+    # compile the model
+    optimiser = keras.optimizers.Adam(learning_rate=learning_rate)
+    model.compile(optimizer=optimiser, loss=error, metrics=["accuracy"])
+
+    # print model overview
+    model.summary()
+
+    return model