I am confused about using these two parameters: epochs & max_history
I have 35 stories. Each story has at least 40 unique lines.
I used--augmentation 0 and epochs 1000 when training the stories.
I used max_history = 10 in bot.py
There was no any problem when I run the interactive learning. However, when I run the bot using debug mode, it went to the wrong path with using keras policy.
Thus, can someone explain that how do these two work?