data/processed/eval_rating.json

{
  "baseline_user_mean": {
    "variant": "baseline_user_mean",
    "n": 1891,
    "rmse": 1.089864042607667,
    "mae": 0.7537457937113194
  },
  "lgbm": {
    "variant": "lgbm",
    "n": 1891,
    "skipped": 0,
    "rmse": 0.6356804161627736,
    "mae": 0.36682990416685235
  },
  "llm_reasoner": {
    "variant": "llm_reasoner",
    "n": 25,
    "rmse": 1.019803902718557,
    "mae": 0.56,
    "avg_latency_sec": 4.5514083671569825
  }
}