How to calculate accuracy during training?

pierluigiferrari / ssd_keras

A Keras port of Single Shot MultiBox Detector

Apache License 2.0

1.86k stars 938 forks source link

How to calculate accuracy during training? #366

Open sachinkmohan opened 4 years ago

sachinkmohan commented 4 years ago

How to calculate accuracy during training?

I tried the below in the code in ssd7_training.ipynb file and didn't work. Any help would be awesome!

model.compile(optimizer=adam, loss=ssd_loss.compute_loss, metrics=['accuracy'])

    * Operating System - Ubuntu 18.04
    * Keras version - 2.2.5
    * TensorFlow version - TF 1.14

stale[bot] commented 3 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

Lucky2593 commented 3 years ago

This is a regression problem, I think that's why we only can get a loss for the problem.

NOTE: You can add the following line :- model.compile(optimizer=sgd,loss=ssd_loss.compute_loss, metrics=[tf.keras.metrics.Accuracy()]) But you will get accuracy as "00+e00".(because it's a regression problem.)