trainer_config.json

{
    "batch_size": 64,
    "learning_rate": 0.0001,
    "weight_decay": 0.01,
    "max_epochs": 60,
    "tasks": {
        "prediction": 0.6,
        "lm": 0.4
    },
    "compile": true,
    "enable_ddp": false,
    "dtype": "float32",
    "num_workers": 16,
    "beta1": 0.9,
    "beta2": 0.95,
    "gradient_accumulation_steps": 1,
    "grad_clip": 1.0,
    "decay_lr": true,
    "log_interval": 10,
    "save_interval": 5,
    "min_lr": 1e-06,
    "warmup_iters": 54
}