Poor performance in Test Set

[HCA97]

all the models performed very badly in round 2 investigate why.

Assumptions:

• [x] #8
• [x] #9
• [x] #10

[HCA97]

It seems like the problem was labeling issues :)

Now our test score is 0.02 points less than our validation score.

https://discourse.aicrowd.com/t/the-submission-needs-japonicus-koreicus-class-not-the-japonicus-koreicus/9139)

[HCA97]

Another reason might https://github.com/HCA97/Mosquito-Classifiction/issues/23#issuecomment-1751823667