Keras SMS Callback

smscallback.py

import os
import tensorflow as tf
from twilio.rest import Client


class SMSCallback(tf.keras.callbacks.Callback):
    """
    SMSCallback
    
    Because you are not always standing in front of your computer waiting for
    the next epoch to be completed. This callback will use your Twilio account
    to send you (or anyone else you want to keep informed) the evolution of
    the training. Your ACCOUNT_SID and AUTH_TOKEN needs to be in your
    environment variables.
    """
    def __init__(self, to, sender, period=1):
        """
        Params
        ------
        
        to: str or list(str)
            Number(s) to send the message to. Format: '+15555555555'
            
        sender: str
            The number from twilio that you want to use. Same format as 'to': '+15555555555'
            
        period: int
            How often to send the SMS
        """
        super(SMSCallback, self).__init__()
        if isinstance(to, str):
            self.to = [to]
        self.sender = sender
        self.client = Client(
            os.environ['ACCOUNT_SID'],
            os.environ['AUTH_TOKEN']
        )
        self.period = period
        
    def on_epoch_end(self, epoch, logs={}):
        if (epoch + 1) % self.period != 0:
            return
        msg = "epoch {epoch} | loss: {loss:.4f}, acc: {acc:.4f}"
        if "val_loss" in logs:
            msg += " val_loss: {val_loss:.4f}, val_acc: {val_acc:.4f}"

        logs['epoch'] = epoch
        for receiver in self.to:
            self.client.messages.create(
                to=receiver,
                from_=self.sender,
                body=msg.format(**logs)
            )