models/label.py

"""Pydantic schemas for key-point matching prediction endpoints"""

from pydantic import BaseModel, Field, ConfigDict
from typing import List, Optional, Dict


class PredictionRequest(BaseModel):
    """Request model for single key-point/argument prediction"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "argument": "Climate change is accelerating due to industrial emissions.",
                "key_point": "Human industry contributes significantly to global warming."
            }
        }
    )

    argument: str = Field(
        ..., min_length=5, max_length=1000,
        description="The argument text to evaluate"
    )
    key_point: str = Field(
        ..., min_length=5, max_length=500,
        description="The key point used for comparison"
    )


class PredictionResponse(BaseModel):
    """Response model for single prediction"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "prediction": 1,
                "confidence": 0.874,
                "label": "apparie",
                "probabilities": {
                    "non_apparie": 0.126,
                    "apparie": 0.874
                }
            }
        }
    )

    prediction: int = Field(..., description="1 = apparie, 0 = non_apparie")
    confidence: float = Field(..., ge=0.0, le=1.0,
                              description="Confidence score of the prediction")
    label: str = Field(..., description="apparie or non_apparie")
    probabilities: Dict[str, float] = Field(
        ..., description="Dictionary of class probabilities"
    )


class BatchPredictionRequest(BaseModel):
    """Request model for batch predictions"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "pairs": [
                    {
                        "argument": "School uniforms limit students' self-expression and creativity",
                        "key_point": "Uniforms restrict personal freedom and individuality"
                    },
                    {
                        "argument": "Renewable energy creates more jobs than fossil fuel industries",
                        "key_point": "Green energy generates employment opportunities"
                    },
                    {
                        "argument": "We should invest more in renewable energy to combat climate change",
                        "key_point": "Capital punishment violates human rights"
                    },
                    {
                        "argument": "Online education provides flexibility and accessibility for all students",
                        "key_point": "Digital learning offers convenient and inclusive education"
                    },
                    {
                        "argument": "Vaccinations are essential for public health and disease prevention",
                        "key_point": "Space exploration leads to scientific discoveries"
                    }
                ]
            }
        }
    )

    pairs: List[PredictionRequest] = Field(
        ..., max_length=100,
        description="List of argument-keypoint pairs (max 100)"
    )


class BatchPredictionResponse(BaseModel):
    """Response model for batch key-point predictions"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "predictions": [
                    {
                        "prediction": 1,
                        "confidence": 0.956,
                        "label": "apparie",
                        "probabilities": {
                            "non_apparie": 0.044,
                            "apparie": 0.956
                        }
                    },
                    {
                        "prediction": 1,
                        "confidence": 0.892,
                        "label": "apparie",
                        "probabilities": {
                            "non_apparie": 0.108,
                            "apparie": 0.892
                        }
                    },
                    {
                        "prediction": 0,
                        "confidence": 0.934,
                        "label": "non_apparie",
                        "probabilities": {
                            "non_apparie": 0.934,
                            "apparie": 0.066
                        }
                    },
                    {
                        "prediction": 1,
                        "confidence": 0.878,
                        "label": "apparie",
                        "probabilities": {
                            "non_apparie": 0.122,
                            "apparie": 0.878
                        }
                    },
                    {
                        "prediction": 0,
                        "confidence": 0.967,
                        "label": "non_apparie",
                        "probabilities": {
                            "non_apparie": 0.967,
                            "apparie": 0.033
                        }
                    }
                ],
                "total_processed": 5,
                "summary": {
                    "total_apparie": 3,
                    "total_non_apparie": 2,
                    "average_confidence": 0.9254
                }
            }
        }
    )

    predictions: List[PredictionResponse]
    total_processed: int = Field(..., description="Number of processed items")
    summary: Dict[str, float] = Field(
        default_factory=dict,
        description="Summary statistics of the batch prediction"
    )


class HealthResponse(BaseModel):
    """Health check model for the API"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "status": "healthy",
                "model_loaded": True,
                "device": "cpu",
                "model_name": "NLP-Debater-Project/destlibert-keypoint-matching",
                "timestamp": "2024-01-01T12:00:00Z"
            }
        }
    )
    
    status: str = Field(..., description="API health status")
    model_loaded: bool = Field(..., description="Whether the model is loaded")
    device: str = Field(..., description="Device used for inference (cpu/cuda)")
    model_name: Optional[str] = Field(None, description="Name of the loaded model")
    timestamp: str = Field(..., description="Timestamp of the health check")


class ModelInfoResponse(BaseModel):
    """Detailed model information response"""
    model_config = ConfigDict(
        json_schema_extra={
            "example": {
                "model_name": "NLP-Debater-Project/destlibert-keypoint-matching",
                "device": "cpu",
                "max_length": 256,
                "num_labels": 2,
                "loaded": True,
                "performance": {
                    "accuracy": 0.9285,
                    "f1_score": 0.8836,
                    "f1_apparie": 0.8113,
                    "f1_non_apparie": 0.9559
                },
                "description": "DistilBERT model for key point - argument semantic matching"
            }
        }
    )

    model_name: str
    device: str
    max_length: int
    num_labels: int
    loaded: bool
    performance: Dict[str, float] = Field(
        ..., description="Model performance metrics"
    )
    description: str = Field(..., description="Model description")