I had sometime today to think about fine-tuning a model based on one of our dataset. It consists of 15k interactions. The training seems to give a very good results. I would like to see how generalised this can be made for other customers. If possible we can have our model scale to maximum of our customer base.

I feel that it is important to give customers best possible outcomes. Every penny they spend should give them max outputs. In order for our customers to get that we should be having best model trained and harnessed in best ways.

pansare