Hi Hanjun,
Please find below the code for DMALA. I tested the performance for bernoulli and categorical models and they seem to match with haoran results.
Note that the below PR also contains the last PR that you have approved (there were some merge issues so I re-included the changes here.)
Hi Hanjun, Please find below the code for DMALA. I tested the performance for bernoulli and categorical models and they seem to match with haoran results. Note that the below PR also contains the last PR that you have approved (there were some merge issues so I re-included the changes here.)