你好我想请教下， 如何将.h5的权重换成.onnx的呢

[tzl1]

以下是我做的一些尝试，但是都失败了：
import keras2onnx,keras
import onnx
from nets.yolo4 import yolo_body
from keras.models import load_model
num_anchors = 9
num_classes = 12
model_path = r'E:/python_pro/test/model_data/last1.h5'
model = yolo_body(Input(shape=(None,None,3)), num_anchors//3, num_classes)
model.load_weights(model_path)
onnx_model = keras2onnx.convert_keras(model, model.name)
temp_model_file = r'E:/python_pro/test/model_data/yolo4.onnx'
onnx.save_model(onnx_model, temp_model_file)

此段代码会报如下错误：

Using TensorFlow backend.

WARNING:tensorflow:From C:\ProgramData\Anaconda3\lib\site-packages\tensorflow\python\framework\op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
The ONNX operator number change on the optimization: 1092 -> 474

[bubbliiiing]

sorry，这个真的难住我了，我没有弄过