training_params.json

{
    "data_path": "md6/autotrain-data",
    "model": "microsoft/mdeberta-v3-base",
    "lr": 2.5e-05,
    "epochs": 1,
    "max_seq_length": 1024,
    "batch_size": 32,
    "warmup_ratio": 0.1,
    "gradient_accumulation": 1,
    "optimizer": "adamw_torch",
    "scheduler": "cosine",
    "weight_decay": 0.2,
    "max_grad_norm": 1.0,
    "seed": 42,
    "train_split": "train",
    "valid_split": "validation",
    "text_column": "autotrain_text",
    "target_column": "autotrain_label",
    "logging_steps": 50,
    "project_name": "md6",
    "auto_find_batch_size": true,
    "mixed_precision": "none",
    "save_total_limit": 1,
    "push_to_hub": true,
    "eval_strategy": "steps",
    "username": "madhurjindal",
    "log": "tensorboard",
    "early_stopping_patience": 3,
    "early_stopping_threshold": 0.01
}