{
    "training": {
        "seed": 777,
        "logging_steps": 100,
        "eval_steps": 200,
        "gradient_accumulation_steps": 1,
        "gradient_clip_value": 1.0,
        "learning_rate": 1e-04,
        "betas": [
            0.8,
            0.9
        ],
        "warmup_ratio": 0.1,
        "batch_size": 32,
        "weight_decay": 0.1,
        "precision": "32-true",
        "strategy": "ddp",
        "gradient_checkpointing": false,
        "num_workers": 1
    },
    "modeling": {
        "parameters": {}
    },
    "codec": {
        "audio_window_size": 192000,
        "sample_rate": 48000,
        "hop_length": 160,
        "minimum_data_sample_rate": 32000,
        "upsample_factors": [
            3,
            2
        ],
        "kernel_sizes": [
            7,
            6
        ]
    },
    "checkpointing": {
        "save_steps": 50,
        "collect_health_stats": false,
        "save_intermediate_generations": true,
        "keep_only_last_n_checkpoints": 5,
        "only_load_model_weights": false
    },
    "train_weighted_datasets": {
        "/path/to/your/vectorized_dataset": 1.0
    },
    "val_weighted_datasets": {
        "/path/to/your/vectorized_dataset": 1.0
    },
    "dataset": {
        "randomize_slider": 0,
        "allowed_languages": [],
        "min_dnsmos_score": 0,
        "min_sample_rate": 0,
        "enable_rlhf_training": false
    }
}