{
  "model_details": {
    "provider": "openai",
    "model_requested": "gpt-5.4",
    "model_for_requests": "gpt-5.4",
    "api_base_url": "https://api.openai.com/v1",
    "chat_completions_endpoint": "https://api.openai.com/v1/chat/completions"
  },
  "cache_padding": {
    "enabled": false,
    "target_shared_prefix_tokens": 0,
    "calibration_shared_prefix_tokens": null,
    "target_prompt_tokens": 0,
    "calibration_prompt_tokens": null,
    "calibration_example_id": null,
    "applied_padding_tokens_estimate": 0,
    "examples_with_padding_applied": 0
  },
  "request_control_summary": {
    "configured": {},
    "attempts_total": 115,
    "attempts_with_control_telemetry": 115,
    "per_control": {
      "reasoning_effort": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "thinking_level": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "effort": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "verbosity": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "prompt_cache_key": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "gemini_cached_content": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      },
      "requesty_auto_cache": {
        "configured_value": null,
        "requested_attempts": 0,
        "sent_attempts": 0,
        "accepted_attempts": 0,
        "rejected_attempts": 0,
        "missing_from_final_request_attempts": 0,
        "acceptance_rate": null,
        "rejected_reasons": {},
        "rejected_example_ids": []
      }
    }
  },
  "usage_metadata_summary": {
    "attempts_total": 115,
    "attempts_with_usage_metadata": 115,
    "attempts_with_cached_token_signals": 115,
    "cached_tokens_total_estimate": 64512,
    "cache_read_tokens_total": 64512,
    "cache_write_tokens_total": 0,
    "cache_token_fields_totals": {
      "usage.prompt_tokens_details.cached_tokens": 64512
    },
    "attempts_with_gemini_cached_content_token_signals": 0,
    "gemini_cached_content_token_count_total": 0,
    "gemini_cached_content_token_fields_totals": {}
  },
  "token_usage_totals": {
    "attempts_total": 115,
    "attempts_with_token_usage": 115,
    "attempts_with_output_tokens": 115,
    "attempts_with_cached_input_tokens": 63,
    "attempts_with_thinking_tokens": 0,
    "input_tokens_total": 160978,
    "cached_input_tokens_total": 64512,
    "non_cached_input_tokens_total": 96466,
    "output_tokens_total": 15187,
    "thinking_tokens_total": 0,
    "output_tokens_definition": "total_tokens - prompt_tokens (or completion_tokens + thinking_tokens fallback)"
  },
  "truth_label_count": 115,
  "prediction_count": 115,
  "evaluated_example_count": 115,
  "calibration_metrics": {
    "available": true,
    "sample_count": 115,
    "bin_count": 10,
    "ece": 0.25565217391304323,
    "mce": 0.74,
    "brier_score": 0.2653234782608696
  },
  "first_prompt_timestamp": "2026-03-16T22:16:19.942954Z",
  "last_prompt_timestamp": "2026-03-16T22:17:48.600453Z",
  "overall_time_seconds": 88.657499,
  "overall_time_human": "1m 29s",
  "accuracy": 0.6869565217391305,
  "macro_precision": 0.44806763285024154,
  "macro_recall": 0.3633333333333333,
  "macro_f1": 0.39952693386582994,
  "per_label": {
    "Clause-final pragmatic marker": {
      "precision": 0.9375,
      "recall": 0.6,
      "f1": 0.7317073170731707,
      "support": 25
    },
    "Clause-medial pragmatic marker": {
      "precision": 0.782608695652174,
      "recall": 0.72,
      "f1": 0.7499999999999999,
      "support": 25
    },
    "Discourse marker": {
      "precision": 0.8125,
      "recall": 0.65,
      "f1": 0.7222222222222223,
      "support": 40
    },
    "Quotative marker": {
      "precision": 1.0,
      "recall": 0.9,
      "f1": 0.9473684210526316,
      "support": 20
    },
    "clause-final pragmatic marker": {
      "precision": 0.0,
      "recall": 0.0,
      "f1": 0.0,
      "support": 0
    },
    "clause-medial pragmatic marker": {
      "precision": 0.0,
      "recall": 0.0,
      "f1": 0.0,
      "support": 0
    },
    "discourse marker": {
      "precision": 0.0,
      "recall": 0.0,
      "f1": 0.0,
      "support": 0
    },
    "quotative marker": {
      "precision": 0.0,
      "recall": 0.0,
      "f1": 0.0,
      "support": 0
    },
    "unclassified": {
      "precision": 0.5,
      "recall": 0.4,
      "f1": 0.4444444444444445,
      "support": 5
    }
  },
  "confusion_matrix": {
    "Clause-final pragmatic marker": {
      "Clause-final pragmatic marker": 15,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 3,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 4,
      "clause-medial pragmatic marker": 1,
      "discourse marker": 2,
      "quotative marker": 0,
      "unclassified": 0
    },
    "Clause-medial pragmatic marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 18,
      "Discourse marker": 2,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 4,
      "discourse marker": 1,
      "quotative marker": 0,
      "unclassified": 0
    },
    "Discourse marker": {
      "Clause-final pragmatic marker": 1,
      "Clause-medial pragmatic marker": 2,
      "Discourse marker": 26,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 3,
      "discourse marker": 6,
      "quotative marker": 0,
      "unclassified": 2
    },
    "Quotative marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 1,
      "Quotative marker": 18,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 1,
      "unclassified": 0
    },
    "clause-final pragmatic marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 0,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 0,
      "unclassified": 0
    },
    "clause-medial pragmatic marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 0,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 0,
      "unclassified": 0
    },
    "discourse marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 0,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 0,
      "unclassified": 0
    },
    "quotative marker": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 0,
      "Discourse marker": 0,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 0,
      "unclassified": 0
    },
    "unclassified": {
      "Clause-final pragmatic marker": 0,
      "Clause-medial pragmatic marker": 3,
      "Discourse marker": 0,
      "Quotative marker": 0,
      "clause-final pragmatic marker": 0,
      "clause-medial pragmatic marker": 0,
      "discourse marker": 0,
      "quotative marker": 0,
      "unclassified": 2
    }
  },
  "labels": [
    "Clause-final pragmatic marker",
    "Clause-medial pragmatic marker",
    "Discourse marker",
    "Quotative marker",
    "clause-final pragmatic marker",
    "clause-medial pragmatic marker",
    "discourse marker",
    "quotative marker",
    "unclassified"
  ],
  "total_examples": 115,
  "label_metrics_available": true,
  "run_config": {
    "task_name": "*like* discourse/pragm",
    "prompt_layout": "standard",
    "task_description": "Classification of instance of *like* into four discourse-pragmatic categories: 1. Quotative marker, 2. Discourse marker, 3. Clause-medial pragmatic marker, 4 . Clause-final pragmatic marker",
    "tags": "*like*;discourse;pragmatics;English;v3"
  },
  "source_input_csv": "",
  "source_output_csv": "data/output/like_interrater__openai__gpt54__2026-03-16-23-16.csv",
  "source_labels_csv": ""
}