dactyl-modernbert-base-finetuned

DACTYL Text Detector

Training Configuration

{
    "training_split": "training",
    "evaluation_split": "testing",
    "results_path": "bce-finetuning-modernbert.csv",
    "num_epochs": 1,
    "model_path": "ShantanuT01/dactyl-modernbert-base-pretrained",
    "tokenizer": "answerdotai/ModernBERT-base",
    "optimizer": "SOTAs",
    "optimizer_type": "libauc",
    "optimizer_args": {
        "lr": 1e-05
    },
    "loss_fn": "tpAUC_KL_Loss",
    "reset_classification_head": true,
    "loss_type": "libauc",
    "loss_fn_args": {
        "data_len": 466005
    },
    "needs_loss_fn_as_parameter": false,
    "save_path": "ShantanuT01/dactyl-modernbert-base-finetuned",
    "training_args": {
        "batch_size": 64,
        "needs_sampler": true,
        "needs_index": true,
        "shuffle": false,
        "sampling_rate": 0.5,
        "apply_sigmoid": true
    },
    "best_model_path": "best-tpauc-model"
}

Results