{
  "seeds": [0, 1, 2],
  "ppo": {
    "steps": 12000,
    "horizon": 1024,
    "gamma": 0.99,
    "learning_rate": 0.0003,
    "update_epochs": 8,
    "clip_eps_sweep": [0.05, 0.2, 0.4],
    "gae_lambda_sweep": [0.8, 0.95, 1.0],
    "value_coef": 0.5,
    "entropy_coef": 0.01
  },
  "sac": {
    "steps": 12000,
    "gamma": 0.99,
    "batch_size": 128,
    "learning_rate": 0.0003,
    "fixed_alpha": 0.2,
    "target_entropy": -1.0,
    "polyak_tau": 0.01
  }
}