Skip to content

ekremcet/Cifar10-Image-classification-with-Tensorflow

Repository files navigation

Cifar10 image classificiation with Tensorflow

ℹ️ Models are designed with 500.000 hidden parameter limit
ℹ️ Achieved accuracy is 71.58% (top1 prediction) with 170 iterations using first model

🔷 Check Trainer.ipynb for more information about models
🔷 McNemar test is used for comparing two models in the evaluation part. You can check the results in evaulator.ipynb