mnist
ml-intern

Built with ML Intern

Small CNN (~225k params) trained from scratch on MNIST for exactly 30 seconds of training time on an Nvidia T4. Adam, lr=1e-3, batch 256, seed 0.

  • Test accuracy: 0.9893 (9893/10000) after 1030 steps in 30.1s
  • Checkpoint: model.pt (torch.save dict with state_dict)
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