Hi,
I try to use TF history and to report the metrics in the end of the training. this is how I try to do it:history = model.fit(.....) logger.report_scalar("accuracy score", "Train", iteration=history.epoch, value=history.history['accuracy']) logger.report_scalar("loss", "Train", iteration=history.epoch, value=history.history['loss'])history is an object ( tf.keras.callbacks.History)
I got en error:TypeError: float() argument must be a string or a number, not 'list'
how I can solve this issue ?
thanks
Hi RoundSeahorse20 ,
According to the error, history.history['accuracy'] and history.history['loss'] are lists containing the values for the accuracy and loss.
You can go over those and report each value:for idx, val in enumerate(history.history['accuracy']): logger.report_scalar("accuracy score", "Train", iteration=idx, value=val)Can this do the trick?
