CIFAR-10 image classifier
ResNet-18 fine-tuned in Google Colab.
Classes: airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck
1 10
Test accuracy: 88.15%
{
"dataset": "CIFAR10" ,
"smoke_test": false ,
"test_accuracy": 0.8815 ,
"validation_accuracy": 0.8905 ,
"winning_run": {
"run_id": "ff235831a9714165862f2b0900a04063" ,
"learning_rate": 0.0003 ,
"validation_accuracy": 0.8905 ,
"best_epoch": 2
} ,
"trials": [
{
"run_id": "a460a30998004a448e417945eeb3791e" ,
"learning_rate": 0.001 ,
"validation_accuracy": 0.849 ,
"best_epoch": 2
} ,
{
"run_id": "ff235831a9714165862f2b0900a04063" ,
"learning_rate": 0.0003 ,
"validation_accuracy": 0.8905 ,
"best_epoch": 2
}
] ,
"seed": 42 ,
"train_samples": 10000 ,
"validation_samples": 2000 ,
"test_samples": 2000 ,
"device": "cuda" ,
"torch_version": "2.11.0+cu130" ,
"submission_branch": "s_109197_103216"
}