File size: 1,500 Bytes
b7ac8b4 475ca58 71c9a9a 475ca58 9ce8001 22085d2 b7ac8b4 e8cce98 9ce8001 b7ac8b4 475ca58 5f5fd58 475ca58 b7ac8b4 475ca58 b7ac8b4 475ca58 b7ac8b4 ee77c85 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
from fastapi import FastAPI, Query
from pydantic import BaseModel
from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
app = FastAPI()
# Load the pre-trained model and tokenizer
model_name = "gyesibiney/Sentiment-review-analysis-roberta-3"
model = AutoModelForSequenceClassification.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)
# Create a sentiment analysis pipeline
sentiment = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)
# Create a dictionary to map sentiment labels to positive and negative strings
sentiment_label_mapping = {
"LABEL_1": "positive",
"LABEL_0": "negative",}
# Define a request body model
class SentimentRequest(BaseModel):
text: str
# Define a response model
class SentimentResponse(BaseModel):
sentiment: str # 1 for positive, 0 for negative
score: float
# Create an endpoint for sentiment analysis with query parameter
@app.get("/sentiment/")
async def analyze_sentiment(text: str = Query(..., description="Input text for sentiment analysis")):
result = sentiment(text)
sentiment_label = result[0]["label"]
sentiment_score = result[0]["score"]
sentiment_value = sentiment_label_mapping.get(sentiment_label, -1) # Default to -1 for unknown labels
return SentimentResponse(sentiment=sentiment_value, score=sentiment_score)
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=8000)
|