Hi Talha Kabakus,

I have been seen you stack overflow issue.

Everything is good.

Please do the training until reach the avg loss less than 0.06. Because that is the good model. Also, increase your dataset to recognize the pattern.