llama-absa-electronics/eval_results.json

{
  "f1_macro": 0.7338,
  "precision": 0.7686,
  "recall": 0.7055,
  "parse_errors": 0,
  "avg_latency_ms": 1024.63
}