Skip to content

Latest commit

 

History

History
59 lines (58 loc) · 10.6 KB

2025-01-11-phi4-instructor-results.md

File metadata and controls

59 lines (58 loc) · 10.6 KB

Benchmark Results (model=phi4, date=2025-01-11)

Domain Response Model Sample Response JSON
document.invoice Invoice 
{
  "invoice_id": null,
  "period_start": null,
  "period_end": null,
  "invoice_issue_date": null,
  "invoice_due_date": null,
  "order_id": null,
  "customer_id": null,
  "issuer": null,
  "issuer_address": null,
  "customer": null,
  "customer_email": null,
  "customer_phone": null,
  "customer_billing_address": null,
  "customer_shipping_address": null,
  "items": null,
  "subtotal": null,
  "tax": null,
  "total": null,
  "currency": null,
  "notes": null
}
document.receipt Receipt 
{
  "receipt_id": null,
  "transaction_date": null,
  "merchant_name": null,
  "merchant_address": null,
  "merchant_phone": null,
  "cashier_name": null,
  "register_number": null,
  "customer_name": "John Doe",
  "customer_id": null,
  "items": [
    {
      "description": "Espresso Coffee",
      "quantity": 2.0,
      "unit_price": 3.5,
      "total_price": 7.0
    },
    {
      "description": "Blueberry Muffin",
      "quantity": 1.0,
      "unit_price": 2.75,
      "total_price": 2.75
    }
  ],
  "subtotal": 9.75,
  "tax": 0.78,
  "total": 11.53,
  "currency": "USD",
  "payment_method": {
    "type": "Credit Card",
    "card_last_4": null,
    "card_type": null
  },
  "discount_amount": null,
  "discount_description": null,
  "tip_amount": 1.0,
  "return_policy": null,
  "barcode": null,
  "additional_charges": null,
  "notes": "Thank you for visiting!",
  "others": {}
}
document.resume Resume 
❌
document.us-drivers-license USDriversLicense 
{
  "issuing_state": "NY",
  "license_number": "123456789",
  "full_name": "John Doe",
  "first_name": "John",
  "middle_name": null,
  "last_name": "Doe",
  "address": {
    "street": "123 Main St",
    "city": "New York",
    "state": "NY",
    "zip_code": "10001"
  },
  "date_of_birth": "1980-05-15",
  "gender": "M",
  "height": "5'11\"",
  "weight": 180.0,
  "eye_color": "Blue",
  "hair_color": "Brown",
  "issue_date": "2022-01-10",
  "expiration_date": "2032-01-10",
  "license_class": "B",
  "donor": true,
  "veteran": false
}
document.utility-bill UtilityBill 
{
  "account_number": "123456789",
  "date_mailed": "2023-09-15",
  "service_for": "John Doe",
  "service_address": {
    "street": "123 Elm Street",
    "city": "Springfield",
    "state": "IL",
    "zip_code": "62704"
  },
  "billing_period_start": "2023-08-01",
  "billing_period_end": "2023-08-31",
  "date_due": "2023-10-15",
  "amount_due": 150.75,
  "previous_balance": 25.0,
  "payment_received": 50.0,
  "current_charges": 175.75,
  "breakdown_of_charges": [
    {
      "description": "Electricity Usage",
      "amount": 100.75,
      "usage": "500 kWh",
      "rate": 0.2
    },
    {
      "description": "Service Fee",
      "amount": 10.0,
      "usage": null,
      "rate": null
    },
    {
      "description": "Taxes",
      "amount": 65.0,
      "usage": null,
      "rate": null
    }
  ],
  "payment_options": [
    "Online Payment",
    "Check by Mail",
    "In-Person"
  ],
  "contact_information": {
    "customer_support": "1-800-555-1234",
    "billing_inquiries": "1-800-555-5678",
    "emergency_services": "911"
  }
}
document.w2-form W2Form 
{
  "control_number": "123456789",
  "ein": "12-3456789",
  "ssn": "987-65-4321",
  "employee_name": "John Doe",
  "employee_address": {
    "street": "123 Main St",
    "city": "Anytown",
    "state": "NY",
    "zip_code": "12345"
  },
  "employer_name": "Acme Corporation",
  "employer_address": {
    "street": "456 Elm St",
    "city": "Othertown",
    "state": "CA",
    "zip_code": "67890"
  },
  "wages_tips_other_compensation": 50000.0,
  "federal_income_tax_withheld": 4000.0,
  "social_security_wages": 49000.0,
  "social_security_tax_withheld": 3036.8,
  "medicare_wages_and_tips": 50000.0,
  "medicare_tax_withheld": 735.0,
  "tax_year": 2022
}
aerospace.remote-sensing RemoteSensing 
{
  "description": "The satellite image depicts a diverse landscape featuring urban development alongside natural features. The central area shows a well-developed residential zone with visible roads and housing structures, indicating a bustling community. Surrounding this urban core are patches of greenery, likely parks or small forested areas, providing a contrast to the built environment. To the east, there is an expansive body of water, possibly a lake or reservoir, bordered by what appears to be a commercial area with several large buildings and parking lots. The western edge of the image reveals agricultural lands, characterized by neatly arranged fields that suggest farming activities. A river meanders through the landscape, connecting various land features and contributing to the region's ecological diversity.",
  "objects": [
    "residential-area",
    "commercial-area",
    "park",
    "lake",
    "farmlands",
    "river"
  ],
  "categories": [
    "residential-area",
    "commercial-area",
    "park",
    "lake",
    "farmlands",
    "river"
  ],
  "is_visible": true
}
healthcare.medical-insurance-card MedicalInsuranceCard 
{
  "provider_service": {
    "provider_service_number": "123456789",
    "precertification_number": null
  },
  "member_information": {
    "member_name": "John Doe",
    "member_id": "987654321",
    "group_number": "ABC123"
  },
  "pharmacy_plan": {
    "rx_bin": "111222333",
    "rx_pcn": "444555666",
    "rx_grp": null,
    "pharmacy_help_desk": "1-800-PHARMACY"
  },
  "insurance_provider": {
    "provider_name": "HealthFirst Insurance Co.",
    "network": "PPO Network"
  },
  "coverage": {
    "office_visit": "In-Network: $20 copay",
    "specialist_visit": "In-Network: $40 copay",
    "urgent_care": "In-Network: $50 visit fee",
    "emergency_room": "In-Network: Deductible applies, then 20% coinsurance",
    "inpatient_hospital": "In-Network: $1,000 deductible per admission"
  }
}
retail.ecommerce-product-caption RetailEcommerceProductCaption 
{
  "description": "The product is a sleek, modern e-reader with a high-resolution display and an ergonomic design for comfortable reading. It features adjustable lighting to reduce eye strain in various environments.",
  "rating": 85,
  "name": "E-Reader Pro",
  "brand": "TechRead",
  "category": "Electronics / E-readers",
  "price": "$199.99",
  "color": "Matte Black"
}
media.tv-news TVNews 
{
  "description": null,
  "chyron": null,
  "network": null,
  "reporters": null
}
media.nfl-game-state NFLGameState 
{
  "description": "The game is currently in progress with Team A leading by a narrow margin.",
  "teams": [
    {
      "name": "Team A",
      "score": 21
    },
    {
      "name": "Team B",
      "score": 19
    }
  ],
  "status": "in_progress",
  "quarter": 3,
  "clock_time": "07:32",
  "possession_team": "Team A",
  "down": "2nd",
  "distance": 5,
  "yard_line": 45,
  "network": "ESPN",
  "is_shown": true
}
media.nba-game-state NBAGameState 
{
  "description": "The game is currently in progress with Team A leading by a narrow margin.",
  "teams": [
    {
      "name": "Team A",
      "score": 102
    },
    {
      "name": "Team B",
      "score": 98
    }
  ],
  "status": "in_progress",
  "quarter": 3,
  "clock_time": "4:32",
  "shot_clock": 14,
  "network": "ESPN",
  "is_shown": true
}