{
  "encoder": "jhu-clsp/mmBERT-base",
  "head_layers": 2,
  "max_len": 3072,
  "head_max_len": 1536,
  "max_prefixes": 6,
  "act_costs": {
    "escalate": 0.5
  },
  "cost_wrong_act": 3.0,
  "amp_dtype": "bf16",
  "model_name": "rl-agent",
  "temperature": [
    1.1609286069869995,
    1.0,
    1.0
  ],
  "temperature_by_options": {},
  "training": {
    "laya_android": {
      "name": "laya-android-v0",
      "base": "multilingual",
      "train_steps": null,
      "seed": 0,
      "epochs": 3,
      "micro_batch": 8,
      "grad_accum": 4,
      "encoder_lr": 2.5e-05,
      "head_lr": 0.0001,
      "loss": "soft-ce",
      "shuffle_options": [
        "choice"
      ],
      "gradient_checkpointing": true
    },
    "best_epoch": 2,
    "train_hours": 2.726,
    "hardware": "NVIDIA GeForce RTX 4090"
  },
  "fine_tuned": true
}