How to decide number of candidates

[alishan2040]

I trained the model on my custom dataset in which each line contains 3 events (e1, e2, e3) (a tri-gram) where I want to predict e3 based on e1 and e2. An example is shown below: After training for 50 epochs I got:

On the predict stage, I kept num_candidates=1, then I got very low F1-score

num_classes in my dataset are 9. If I increase, num_candidates, divide by zero error occurred. What is the best approach to choose num_candidates for given dataset?

Thanks

[shoaib-intro]

If I try to address this increasing candidate size will increase accuracy and will be less sensitive to anomalies vise versa. threshold here is window size.