suramadu08/trash-classification-vit

vit-trash

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on an garythung/trashnet dataset.

Training and evaluation data

DatasetDict({
    train: Dataset({
        features: ['image', 'label'],
        num_rows: 3537
    })
    validation: Dataset({
        features: ['image', 'label'],
        num_rows: 758
    })
    test: Dataset({
        features: ['image', 'label'],
        num_rows: 759
    })
})

Training hyperparameters

TrainingArguments(
    output_dir="output-models",
    save_total_limit=2,
    report_to="wandb",
    save_strategy="epoch",
    evaluation_strategy="epoch",
    learning_rate=2e-5,
    per_device_train_batch_size=15,
    per_device_eval_batch_size=10,
    num_train_epochs=3,
    weight_decay=0.01,
    load_best_model_at_end=True,
    logging_dir="logs",
    remove_unused_columns=False,
)

Training results

Framework versions

• Transformers 4.41.0
• TensorFlow 2.15.0
• Datasets 2.19.1
• Tokenizers 0.19.1