Mengembalikan respons prediksi terstruktur

Menerapkan AI ke Produksi dengan FastAPI

Matt Eckerle

Software and Data Engineering Leader

Tantangan dalam menerapkan model

Diagram alur yang menampilkan prediksi ML dengan FastAPI

Terima data input dengan benar
Validasi data masuk dan tangani error
Lakukan prediksi
Kembalikan respons yang terstruktur baik

Mendefinisikan struktur permintaan

from pydantic import BaseModel

class PredictionRequest(BaseModel):
    text: str

class PredictionResponse(BaseModel):

    text: str

    sentiment: str

    confidence: float

Membuat endpoint prediksi

@app.post("/predict")
def predict_sentiment(request: PredictionRequest):
    if sentiment_model is None:
        raise HTTPException(
            status_code=503,
            detail="Model not loaded"
        )

    result = sentiment_model(request.text)

    return PredictionResponse(
            text=request.text,
            sentiment=result[0]["label"],
            confidence=result[0]["score"]
        )

Input JSON:

{"text": "This movie was fantastic!"}

Respons:

{
    "text": "This movie was fantastic!",
    "sentiment": "POSITIVE",
    "confidence": 0.95
}

Penanganan error

try:
  result = sentiment_model(request.text)
  return PredictionResponse(
        text=request.text,
        sentiment=result[0]["label"],
        confidence=result[0]["score"]
    )
except Exception:
  raise HTTPException(
        status_code=500,
        detail="Prediction failed"
    )

Respons saat model gagal memprediksi

{
    "detail": "Prediction failed",
    "status_code": 500
}

Menguji endpoint

# Example request
import requests

response = requests.post(
    "http://localhost:8000/predict",
    json={"text": "Great product!"}
)
print(response.json())

{
    "text": "Great product!",
    "sentiment": "POSITIVE",
    "confidence": 0.998
}

Ayo berlatih!

Menerapkan AI ke Produksi dengan FastAPI