abidlabs
/

mnist-1min

Model card Files Files and versions

MNIST CNN — trained from scratch in 1 minute

Model: small CNN trained from random init (0.90M params)
Data: ylecun/mnist (train split), evaluated on the test split
Method: Adam lr=1e-3, batch size 256, time-bounded training (60s)
Final test accuracy: 99.47%

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support