epoch: 0 training_loss 0.4078448025882244 test_loss: 0.3148078680038452 test_wrong: 0.1578125
epoch: 1 training_loss 0.2686599436402321 test_loss: 0.2330552101135254 test_wrong: 0.0734375
epoch: 2 training_loss 0.21095163606107234 test_loss: 0.20937373638153076 test_wrong: 0.084375
epoch: 3 training_loss 0.18619711801409722 test_loss: 0.21208367347717286 test_wrong: 0.096875
epoch: 4 training_loss 0.17209375232458116 test_loss: 0.20416505336761476 test_wrong: 0.090625
