metadata
datasets:
- ylecun/mnist
- zalando-datasets/fashion_mnist
language:
- en
- ru
metrics:
- accuracy
- confusion_matrix
How to run
Download mnist dataset: python dataset_downloader.py -f "MNIST" -d "ylecun/mnist"
Download fashion mnist dataset: python dataset_downloader.py -f "Fashion-MNIST" -d "zalando-datasets/fashion_mnist"
Train: python mnist.py --mode train --dataset "./Fashion-MNIST/train.csv" --model "./models/Fashion-MNIST.pth" --epochs 8 --batch-size 64 --lr 0.001
Inference: python mnist.py --mode inference --input "./Fashion-MNIST/test.csv" --model "./models/Fashion-MNIST.pth" --output "./Fashion-MNIST/predictions.csv" --batch-size 64
Evaluate: python evaluate.py --ground-truth "./Fashion-MNIST/test.csv" --predictions "./Fashion-MNIST/predictions.csv"