訓練模型一直跑不出結果

results = {}
for i, (use_preact, bz) in enumerate(itertools.product(PRE_ACTIVATE, BATCH_SIZE)):
    print("Numbers of exp: %i, use_preact: %s, batch_size: %i" % (i, use_preact, bz))
    model = build_mlp(input_shape=x_train.shape[1:], pre_activate=use_preact)
    model.summary()
    optimizer = keras.optimizers.Adam(lr=LEARNING_RATE)
    model.compile(loss="categorical_crossentropy", metrics=["accuracy"], optimizer=optimizer)

    model.fit(x_train, y_train, 
              epochs=EPOCHS, 
              batch_size=bz, 
              validation_data=(x_test, y_test), 
              verbose=0,
              shuffle=True)
    
    # Collect results
    exp_name_tag = ("exp-%s" % (i))
    results[exp_name_tag] = {'train-loss': model.history.history["loss"],
                             'valid-loss': model.history.history["val_loss"],
                             'train-acc': model.history.history["acc"],
                             'valid-acc': model.history.history["val_acc"]}

results = {}
for i, (use_preact, bz) in enumerate(itertools.product(PRE_ACTIVATE, BATCH_SIZE)):
    print("Numbers of exp: %i, use_preact: %s, batch_size: %i" % (i, use_preact, bz))
    model = build_mlp(input_shape=x_train.shape[1:], pre_activate=use_preact)
    model.summary()
    optimizer = keras.optimizers.Adam(lr=LEARNING_RATE)
    model.compile(loss="categorical_crossentropy", metrics=["accuracy"], optimizer=optimizer)

    model.fit(x_train, y_train, 
              epochs=EPOCHS, 
              batch_size=bz, 
              validation_data=(x_test, y_test), 
              verbose=0,
              shuffle=True)
    
    # Collect results
    exp_name_tag = ("exp-%s" % (i))
    results[exp_name_tag] = {'train-loss': model.history.history["loss"],
                             'valid-loss': model.history.history["val_loss"],
                             'train-acc': model.history.history["acc"],
                             'valid-acc': model.history.history["val_acc"]}

訓練模型一直跑不出結果

回答列表