#  [74th place] First Medal - How I missed the Silver Medal

Thank you to the organizers for this interesting competition!
Congrats to all the winners and participants.
In fact, I haven't tried a lot.
We realized the way too late.
However it can be helpful for someone, so simply write it down.
(I did other experiments - pretrained other models, optimizers, etc. - but I'm going to write down the main issues.)
EfficientNet-B0, EfficientNet-B1, EfficientNet-B2
It was good to do simple experiments, but it was hard to exceed 0.930 lb.
EfficientNet-B7
And then I trained b7.
It had the best performance on the about epoch 30 and achieved 0.933 lb.
(I haven't tested many epochs due to limited GPU resources.)
EfficientNet-B6
Learned just before the close of the competition and achieved 0.930 lb at about 28 epoch.
Agumentation
flip augmentation
rotate +- 10 (just B6)
cutmix (our teammate - B3, @
[kani23](https://www.kaggle.com/kani23)
)
When I applied rotate and cutmix, the validation score was a little higher.
like this:
[RESULT]: Train. Epoch: 21, summary_loss: 0.82028, final_score: 0.87635, time: 4017.75592
[RESULT]: Val. Epoch: 21, summary_loss: 0.75268, final_score: 0.90341, time: 325.94613
TTA
Above 0.930 lb, it didn't work in public lb.
but in private, Actually, there was a very big difference.
EfficientNet-B7 + no TTA :  0.932 lb / (private 0.916)
EfficientNet-B7 + TTA      :  0.932 lb / (private 0.920)
Unfortunately I ensemble with the results of not applying TTA.
So, it didn't score better than my single model.
😭
![](assets/ee4e88701cc4943f46459650.png)
Anyway, I learned a lot and I will try harder in the future.
Thanks!
p.s. Thanks to our teammate @
[kani23](https://www.kaggle.com/kani23)
,
p.s.2.
[](https://www.kaggle.com/loveall)
[@loveall](/loveall)
- You can not use our results for your paper job.
because you have contributed nothing and please don't take someone else's kernel anymore.