Retinanet+keras train their own data set to tread on the pit

1. After training the model , Run the training code , Report errors ：ValueError: not enough values to unpack (expected 3, got 2)
Online search replies are opencv Version issue for .
The current period 2020 year 1 month 8 Japan 09:36:30, The version I use is ：
opencv-contrib-python 4.1.2.30
opencv-python 4.1.2.30
resolvent ：
Version independent .
In the original git Connecting Keras RetinaNet There are also descriptions .
principle ： The trained model needs to be transformed into inference Model
Code ： Insert a code chip here

# Running directly from the repository:
keras_retinanet/bin/convert_model.py /path/to/training/model.h5 /path/to/save/inference/model.h5
# Using the installed script:
retinanet-convert-model /path/to/training/model.h5 /path/to/save/inference/model.h5

Of course , I personally think this way , Too complicated .
Attached below are for personal use modeltest.py

import keras
from keras_retinanet import models
from keras_retinanet.utils.image import read_image_bgr, preprocess_image, resize_image
from keras_retinanet.utils.visualization import draw_box, draw_caption
from keras_retinanet.utils.colors import label_color

import matplotlib.pyplot as plt
import cv2
import os

# Whether to use gpu
os.environ["CUDA_VISIBLE_DEVICES"] = "0"
import numpy as np
import time

import tensorflow as tf


def get_session():
    config = tf.ConfigProto()
    config.gpu_options.allow_growth = True
    return tf.Session(config=config)


#  Set up tensorflow session  by Keras  Back end 
keras.backend.tensorflow_backend.set_session(get_session())
#  The converted inferred model address 
# model_path = os.path.join('..', 'snapshots', 'predict.h5')
model_path = 'E:\\winebottle\\keras-retinanet-master\\keras_retinanet\\bin\\snapshots1\\resnet50_csv_19.h5'
#  Load model 
model = models.load_model(model_path, backbone_name='resnet50')
# solve ValueError: not enough values to unpack (expected 3, got 2)
model = models.convert_model(model)
#  establish ID Mapping dictionary with category 
labels_to_names = {  0：‘ Category 1’}
#  Load the image to be detected 
path = '/home/zbb/keras-retinanet/CSV/test/50.JPG'

save_path = ' Saved path '
image_names = sorted(os.listdir(path))
for image_path in image_names:
    image = read_image_bgr(path + image_path)
    print(path + image_path)
    # copy To another object and to RGB file 
    draw = image.copy()
    draw = cv2.cvtColor(draw, cv2.COLOR_BGR2RGB)
    #  Image preprocessing 
    image = preprocess_image(image)
    image, scale = resize_image(image)
    #  Model to predict 
    start = time.time()


    boxes, scores, labels = model.predict_on_batch(np.expand_dims(image, axis=0))
    print("processing time: ", time.time() - start)

    #  Correction ratio 
    boxes /= scale
    #  Visual display of target detection 
    for box, score, label in zip(boxes[0], scores[0], labels[0]):
        #  Set the lowest threshold of prediction score 
        if score < 0.75:
            break
        color = label_color(label)
        print(color)
        b = box.astype(int)
        draw_box(draw, b, color=color)
        caption = "{} {:.3f}".format(labels_to_names[label], score)
        draw_caption(draw, b, caption)
    #  Pictures show 
    plt.figure(figsize=(15, 15))
    plt.axis('off')
    plt.imshow(draw)
    plt.savefig(save_path + image_path, format='JPG', transparent=True, pad_inches=0, dpi=300, bbox_inches='tight')