{"_id": {"$oid": "69d4eb12b877594bc5718d16"}, "test_id": "T0807", "benchmark": "magazine_pages", "date": "2026-03-23", "tags": {"century": [20], "document-type": ["newspaper-page"], "language": ["en"], "layout": ["prose", "columns"], "script": ["latin"], "task": ["document-understanding"], "writing": ["printed"]}, "contributors": ["lea_kasper", "sorin_marti", "arno_bosse"], "hidden": false, "config": {"provider": "mistral", "model": "mistral-large-2512", "dataclass": "MagazinePage", "temperature": 0.0, "role_description": "You are a historian with expertise in analyzing and summarizing historical magazine pages", "prompt_file": "prompt.txt", "legacy_test": false}, "prompt": "Extract all advertisements and return their bounding boxes.\nThe original size of the page is {width} x {height} pixels.", "results": {"text": "{\"advertisements\":[{\"box\":[100.0,200.0,800.0,600.0]},{\"box\":[900.0,150.0,1600.0,500.0]},{\"box\":[1700.0,300.0,2300.0,800.0]},{\"box\":[100.0,700.0,1200.0,1200.0]},{\"box\":[1300.0,900.0,2000.0,1400.0]},{\"box\":[200.0,1500.0,1000.0,2000.0]},{\"box\":[1100.0,1600.0,2200.0,2200.0]},{\"box\":[50.0,2300.0,700.0,2800.0]},{\"box\":[800.0,2400.0,1800.0,2900.0]},{\"box\":[1900.0,2500.0,2400.0,3200.0]},{\"box\":[100.0,3000.0,1000.0,3400.0]}]}", "model": "mistral-large-2512", "provider": "mistral", "finish_reason": "stop", "usage": {"input_tokens": 254, "output_tokens": 347, "total_tokens": 601, "input_cost_usd": 0.000127, "output_cost_usd": 0.0005204999999999999, "estimated_cost_usd": 0.0006474999999999999}, "duration": 5.642757892608643, "timestamp": "2026-03-23T22:33:32.133487", "parsed": {"advertisements": [{"box": [100, 200, 800, 600]}, {"box": [900, 150, 1600, 500]}, {"box": [1700, 300, 2300, 800]}, {"box": [100, 700, 1200, 1200]}, {"box": [1300, 900, 2000, 1400]}, {"box": [200, 1500, 1000, 2000]}, {"box": [1100, 1600, 2200, 2200]}, {"box": [50, 2300, 700, 2800]}, {"box": [800, 2400, 1800, 2900]}, {"box": [1900, 2500, 2400, 3200]}, {"box": [100, 3000, 1000, 3400]}]}, "conversation_id": "df81daec-b389-465d-ba76-bf687a120016", "score": {"f1": 0.0, "precision": 0.0, "recall": 0.0, "mean_iou": 0.0, "true_positives": 0, "false_positives": 11, "false_negatives": 1, "iou_threshold": 0.5}, "raw_response": "{\"id\":\"40ebca7ecbf747559932c2e0e97484ed\",\"object\":\"chat.completion\",\"model\":\"mistral-large-2512\",\"usage\":{\"prompt_tokens\":254,\"completion_tokens\":347,\"total_tokens\":601,\"prompt_tokens_details\":{\"cached_tokens\":228}},\"created\":1774301607,\"choices\":[{\"index\":0,\"message\":{\"role\":\"assistant\",\"content\":\"{\\n  \\\"advertisements\\\": [\\n    {\\n      \\\"box\\\": [100, 200, 800, 600]\\n    },\\n    {\\n      \\\"box\\\": [900, 150, 1600, 500]\\n    },\\n    {\\n      \\\"box\\\": [1700, 300, 2300, 800]\\n    },\\n    {\\n      \\\"box\\\": [100, 700, 1200, 1200]\\n    },\\n    {\\n      \\\"box\\\": [1300, 900, 2000, 1400]\\n    },\\n    {\\n      \\\"box\\\": [200, 1500, 1000, 2000]\\n    },\\n    {\\n      \\\"box\\\": [1100, 1600, 2200, 2200]\\n    },\\n    {\\n      \\\"box\\\": [50, 2300, 700, 2800]\\n    },\\n    {\\n      \\\"box\\\": [800, 2400, 1800, 2900]\\n    },\\n    {\\n      \\\"box\\\": [1900, 2500, 2400, 3200]\\n    },\\n    {\\n      \\\"box\\\": [100, 3000, 1000, 3400]\\n    }\\n  ]\\n}\",\"tool_calls\":null,\"prefix\":false},\"finish_reason\":\"stop\"}]}"}, "scoring": {"f1": 0.021, "precision": 0.016, "recall": 0.043, "mean_iou": 0.062, "true_positives": 7, "false_positives": 450, "false_negatives": 119, "num_pages": 46, "cost_summary": {"total_input_tokens": 11684, "total_output_tokens": 14578, "total_tokens": 26262, "input_cost_usd": 0.005842000000000004, "output_cost_usd": 0.021866999999999994, "total_cost_usd": 0.02770899999999999}}, "normalized_score": 2.1}