Add application file

app.py

@@ -1,25 +1,40 @@
 from fastapi import FastAPI, Query
-from transformers import pipeline
-
-# Initialize the sentiment analysis pipeline
-pipe = pipeline("text-classification", model="cardiffnlp/twitter-roberta-base-sentiment")
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+from scipy.special import softmax
+from typing import Dict
 
 app = FastAPI()
 
-@app.get('/')
-def home():
-    return {"message": "Welcome to the Sentiment Analysis API"}
+# Load model and tokenizer
+roberta = "cardiffnlp/twitter-roberta-base-sentiment"
+model = AutoModelForSequenceClassification.from_pretrained(roberta)
+tokenizer = AutoTokenizer.from_pretrained(roberta)
+
+labels = ['Negative', 'Neutral', 'Positive']
+
+def preprocess_tweet(tweet: str) -> str:
+    tweet_words = []
+    for word in tweet.split(' '):
+        if word.startswith('@') and len(word) > 1:
+            word = '@user'
+        elif word.startswith('http'):
+            word = "http"
+        tweet_words.append(word)
+    return " ".join(tweet_words)
+
+def analyze_sentiment(tweet: str) -> Dict[str, float]:
+    tweet_proc = preprocess_tweet(tweet)
+    encoded_tweet = tokenizer(tweet_proc, return_tensors='pt')
+    output = model(**encoded_tweet)
+    scores = output[0][0].detach().numpy()
+    scores = softmax(scores)
+    return {labels[i]: float(scores[i]) for i in range(len(scores))}
 
-@app.get('/analyze')
-def analyze(tweet: str = Query(..., description="The tweet to analyze")):
-    # Use the pipeline to analyze the sentiment of the tweet
-    results = pipe(tweet)
-    
-    # Format the output to include labels and their corresponding scores
-    sentiment_scores = {result['label']: result['score'] for result in results}
-    
-    return sentiment_scores
+@app.get("/analyze-sentiment/")
+async def analyze_sentiment_endpoint(tweet: str = Query(..., description="The tweet to analyze")):
+    sentiment_scores = analyze_sentiment(tweet)
+    return {"sentiment": sentiment_scores}
 
-# To run this app on Hugging Face Spaces:
-# 1. Save this code as `app.py`.
-# 2. Deploy it to a new Hugging Face Space, and the Space will automatically start your FastAPI app.
+# To run the application:
+# 1. Save this code as `main.py`.
+# 2. Run the command `uvicorn main:app --reload`.