about train epoch/ train iter

[yang-yee]

Hello author, I found that the for epoch in range (1000000) in the train source code refers to a total of 1M iterations of the model? May I ask if the total number of iterations needs to be modified here? Can I make it smaller? Because it runs slowly on our server, my batch size is 6. How much would be suitable if it were reduced?