# Reading accuracy in Keras from the saved model

Yesterday I trained the model in Keras all night, but when I tried to see the results, my computer was not working. When I turned it ON, then I saw 'model.h5' file built already, but I couldn't see the results of accuracy of my training and testing. I saved the model with 'model.save', but how can I now see the accuracy of the model only from that saved model?

• What type of problem are you dealing with here ? Classification or Regression ? You can not read the accuracy of the model you will have to find it out. Do you have a test set or validation set ? – Kaustubh Jul 4 '18 at 7:53

According to the official Keras website, you have to use:

keras.models.load_model(filepath)


Example:

model = load_model('my_model.h5')


This will load your saved H5 model to 'model' and then you can try:

model.evaluate(x,y)


To return the loss value & metrics values for the model in test mode.