16f50016cabab22efde6f183c89efe0a11b0c683,src/RMDL_Image.py,,image_classifciation,#,22

Before Change

def image_classifciation(X_train, y_train, X_test, y_test, batch_size, shape, sparse_categorical, Random_Deep,
                            n_epochs):
    y_proba = []
    model = []
    score = []
    history_ = []
    number_of_classes = np.max(y_train)+1
    //X_train, y_train, X_test, y_test, shape, number_of_classes = Image_Data_load.load(Data_Image)
    i =0
    while i < Random_Deep[0]:
        try:
            print("MNIST DNN ", i, "\n")
            model.append(Sequential())
            model[i], model_tmp = BuildModel.buildModel_DNN_image(shape, number_of_classes, 0)
            filepath = "weights\weights_DNN_" + str(i) + ".hdf5"
            checkpoint = ModelCheckpoint(filepath, monitor="val_acc", verbose=1, save_best_only=True,
                                         mode="max")
            callbacks_list = [checkpoint]

            history = model[i].fit(X_train, y_train, validation_data=(X_test, y_test),
                                   epochs=n_epochs[0], batch_size=batch_size, callbacks=callbacks_list,
                                   verbose=2)
            history_.append(history)
            model_tmp.load_weights(filepath)

            if sparse_categorical == 0:
                model_tmp.compile(loss="sparse_categorical_crossentropy",
                                  optimizer="adam",
                                  metrics=["accuracy"])
            else:
                model_tmp.compile(loss="categorical_crossentropy",
                                  optimizer="adam",
                                  metrics=["accuracy"])

            y_pr = model_tmp.predict_classes(X_test, batch_size=batch_size)
            y_proba.append(np.array(y_pr))
            score.append(accuracy_score(y_test, y_pr))
            i = i + 1
        except:
            print("Error in model ", i, "   try to re-generate an other model")

    i =0
    while i < Random_Deep[1]:
        try:
            print("RNN ", i, "\n")
            model.append(Sequential())
            model[i], model_tmp = BuildModel.Image_model_RNN(number_of_classes, shape)

            filepath = "weights\weights_RNN_" + str(i) + ".hdf5"
            checkpoint = ModelCheckpoint(filepath, monitor="val_acc", verbose=1, save_best_only=True,

After Change

        y_proba.append(np.array(y_pr))
        score.append(accuracy_score(y_test, y_pr))
        i = i+1
        del model_tmp
        del model_CNN
        gc.collect()

Instances