我是ML和AI的新手，我试图建立一个阅读文本图像的模型，我有点迷失在这里。这是我用来建立模型的代码，但我不能。

import os
import random
import numpy as np
import tensorflow as tf
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense
from tensorflow.keras.models import Sequential

# Defines the path to your images folder
folder_path = r"C:\Users\cuell.DESKTOP-1DJM07S\Desktop\MLTU\train"

# Defines the characters that can appear in the images
characters = "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"

# Define the number of images to use for training
num_train_images = 85

# Define the size of the images
image_width, image_height, num_channels = 180, 40, 3

# Defines the number of characters in each image
num_characters = 6

# Defines the batch size and number of epochs to use for training
batch_size, num_epochs = 10, 1

# Load the images and their labels
def load_images(folder_path):
    images, labels = [], []
    for filename in os.listdir(folder_path):
        if filename.endswith(".jpg"):
            image_path = os.path.join(folder_path, filename)
            image = tf.io.decode_png(tf.io.read_file(image_path), channels=num_channels)
            label = filename.split(".")[0]
            images.append(image)
            labels.append(label)
    return images, labels

images, labels = load_images(folder_path)
# Convert the images and labels to numpy arrays
def convert_to_arrays(images, labels):
    images_array = np.array(images) / 255.0
    labels_array = np.zeros((num_train_images, num_characters, len(characters)))
    for i, label in enumerate(labels):
        for j, char in enumerate(label):
            labels_array[i, j, characters.index(char)] = 1
    return images_array, labels_array

images_array, labels_array = convert_to_arrays(images, labels)

# Shuffle the images and labels
def shuffle_arrays(images_array, labels_array):
    permutation = np.random.permutation(len(images_array))
    return images_array[permutation], labels_array[permutation]

images_array, labels_array = shuffle_arrays(images_array, labels_array)

# Split the images and labels into training and validation sets
def split_arrays(images_array, labels_array):
    num_train = int(num_train_images * 0.8)
    train_images, train_labels = images_array[:num_train], labels_array[:num_train]
    val_images, val_labels = images_array[num_train:], labels_array[num_train:]
    return train_images, train_labels, val_images, val_labels

train_images, train_labels, val_images, val_labels = split_arrays(images_array, labels_array)

# Define the TensorFlow model
model = Sequential([
    Conv2D(32, (3, 3), activation="relu", input_shape=(40, 180, 3)),
    MaxPooling2D(pool_size=(2, 2)),
    Conv2D(64, (3, 3), activation="relu"),
    MaxPooling2D(pool_size=(2, 2)),
    Flatten(),
    Dense(128, activation="relu"),
    Dense(num_characters * len(characters), activation="relu")
])

# Compile the TensorFlow model
model.compile(optimizer="adam", loss="categorical_crossentropy", metrics=["accuracy"])

# Train the TensorFlow model
model.fit(train_images, train_labels, batch_size=batch_size, epochs=num_epochs, validation_data=(val_images, val_labels))

#model.save("text_model.h5")

当我点击www.example.com时model.fit，我得到了ValueError的答案：形状（无，6，36）和（无，216）是不兼容的，我的猜测是，我做错了什么，对分裂的20%的测试和培训，和他们的维度变化，但我仍然不确定。
我期待着我的第一个文本阅读模型运行。：）