Jesse Moore

Results 1 comments of Jesse Moore

This seems a little odd. The readme explicitly states that the max_seq_length of the data prep and the training steps don't need to be identical. > Longer sequences are disproportionately...