Keras LSTM Batch Size and model.fit()

Question

I am training an LSTM in Keras. As per documentation, my training data and labels have shape (20, 20, 1) representing 20 samples with 20 time steps and one feature. When I use model.fit() to train my model, do I need to specify batch size or will all 20 samples be sent as one batch by default?

Answer 1

According to Keras's fit documentation

batch_size Integer or NULL. Number of samples per gradient update. If unspecified, batch_size will default to 32 .