batch parameters name in distil_prediction_layer_from and distil_intermediate_layers_from

[eboraks]

Hello there,

I recently worked with the FARMReader to distill a model, and I noticed that the two distill methods have different batch parameter names.

distil_intermediate_layers_from the parameter is called batch_size and distil_prediction_layer_from uses student_batch_size. According to the documentation, both methods should use student_batch_size.

[julian-risch]

Hello @eboraks thank you for notifying us about that. I agree that we should rename the parameter or adjust the documentation. I created an issue to follow up on this here: #3771