Convert custom Yolov4 .weight file to Tensorflow

Question

I have trained my custom object using Yolov4 and I have the following file :

yolov4-custom_best.weights
yolov4-custom.cfg
obj.names

'obj.names' has the names of the classes of the custom object.

In order to deploy my custom object detector to a web application I need the files in a tensorflow model format (eg : object_detector.h5)

Can someone help?

Answer 1

To get the model files in .h5 format you have to save the model. To save the model you can use

# The '.h5' extension indicates that the model should be saved to HDF5.
model.save('my_model.h5')

You can also save the model by using tf.keras.models.save_model()

tf.keras.models.save_model(model,filepath,save_format='h5')

For more details, please refer to this documentation. Thank You!