Closed dmumtaz closed 1 year ago
For validation, we provide curves for SRCC, PLCC, KRCC, and RMSE computation. They are metrics to know whether the model is overfitting. As validation is only conducted once an epoch, we believe the metrics are enough for monitoring the training and loss curves are not so important. You can also simply copy the related code from "finetune_epoch" to "inference_set" function to monitor them, with wandb apis.
Similar to the training loss, can you please share how to visualize the validation loss V/S epoch graph used for checking the performance and whether the model is overfitting?