{
  "optimizer": "lamb",
  "scheduler": "linear_warmup",
  "learning_rate": 0.0001,
  "batch_size": 24,
  "epochs": 10,
  "seed": 3407,
  "status": "default recipe; not a completed run"
}