Upload backend/app/main copy.py with huggingface_hub

backend/app/main copy.py

@@ -0,0 +1,148 @@
+from flask import Flask, request, jsonify
+from flask_cors import CORS
+import torch
+from transformers import BertTokenizer
+from train import YelpReviewClassifier
+
+# Initialize Flask app
+app = Flask(__name__)
+CORS(app)  # Enable CORS for cross-origin requests
+
+# Load Model
+device = "cuda" if torch.cuda.is_available() else "cpu"
+print(f"Using device: {device}")
+
+model = YelpReviewClassifier().to(device)
+model.load_state_dict(torch.load("models/model.pth", map_location=device))
+model.eval()
+
+tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
+
+@app.route("/")
+def home():
+    return jsonify({"message": "Welcome to Yelp Review AI Predictor!"})
+
+@app.route("/predict", methods=["POST"])
+def predict_rating():
+    try:
+        # Get JSON data from request
+        data = request.get_json()
+        
+        if not data:
+            return jsonify({"error": "No input data provided"}), 400
+
+        text = data.get("text", "").strip()
+        useful = data.get("useful", 0.0)
+
+        # Ensure text is provided
+        if not text:
+            return jsonify({"error": "Missing 'text' field"}), 400
+        
+        # Ensure useful is a number
+        try:
+            useful = float(useful)
+        except ValueError:
+            return jsonify({"error": "'useful' must be a numeric value"}), 400
+
+        # Tokenize input
+        tokens = tokenizer(
+            text, truncation=True, padding="max_length", max_length=256, return_tensors="pt"
+        )
+        input_ids = tokens["input_ids"].to(device)
+        attention_mask = tokens["attention_mask"].to(device)
+        useful_tensor = torch.tensor([useful], dtype=torch.float32).to(device)
+
+        # Make prediction
+        with torch.no_grad():
+            output = model(input_ids, attention_mask, useful_tensor)
+            predicted_stars = torch.argmax(output, dim=1).item() + 1  # Convert to 1-5 stars
+
+        # **Return predicted stars & usefulness score**
+        return jsonify({
+            "predicted_stars": predicted_stars,
+            "usefulness_score": useful  # ✅ Add usefulness score in response
+        })
+
+    except Exception as e:
+        return jsonify({"error": str(e)}), 500
+
+    try:
+        # Get JSON data from request
+        data = request.get_json()
+        
+        if not data:
+            return jsonify({"error": "No input data provided"}), 400
+
+        text = data.get("text", "").strip()
+        useful = data.get("useful", 0.0)
+
+        # Ensure text is provided
+        if not text:
+            return jsonify({"error": "Missing 'text' field"}), 400
+        
+        # Ensure useful is a number
+        try:
+            useful = float(useful)
+            if not (0 <= useful <= 1):  # Ensure it's between 0 and 1
+                return jsonify({"error": "'useful' must be between 0 and 1"}), 400
+        except ValueError:
+            return jsonify({"error": "'useful' must be a numeric value"}), 400
+
+        # Tokenize input
+        tokens = tokenizer(
+            text, truncation=True, padding="max_length", max_length=256, return_tensors="pt"
+        )
+        input_ids = tokens["input_ids"].to(device)
+        attention_mask = tokens["attention_mask"].to(device)
+        useful_tensor = torch.tensor([useful], dtype=torch.float32).to(device)
+
+        # Make prediction
+        with torch.no_grad():
+            output = model(input_ids, attention_mask, useful_tensor)
+            predicted_stars = torch.argmax(output, dim=1).item() + 1  # Convert to 1-5 stars
+
+        return jsonify({"predicted_stars": predicted_stars})
+
+    except Exception as e:
+        return jsonify({"error": str(e)}), 500
+
+    try:
+        # Get JSON data from request
+        data = request.get_json()
+        
+        if not data:
+            return jsonify({"error": "No input data provided"}), 400
+
+        text = data.get("text", "").strip()
+        useful = data.get("useful", 0.0)
+
+        # Ensure text is provided
+        if not text:
+            return jsonify({"error": "Missing 'text' field"}), 400
+        
+        # Ensure useful is a number
+        try:
+            useful = float(useful)
+        except ValueError:
+            return jsonify({"error": "'useful' must be a numeric value"}), 400
+
+        # Tokenize input
+        tokens = tokenizer(
+            text, truncation=True, padding="max_length", max_length=256, return_tensors="pt"
+        )
+        input_ids = tokens["input_ids"].to(device)
+        attention_mask = tokens["attention_mask"].to(device)
+        useful_tensor = torch.tensor([useful], dtype=torch.float32).to(device)
+
+        # Make prediction
+        with torch.no_grad():
+            output = model(input_ids, attention_mask, useful_tensor)
+            predicted_stars = torch.argmax(output, dim=1).item() + 1  # Convert to 1-5 stars
+
+        return jsonify({"predicted_stars": predicted_stars})
+
+    except Exception as e:
+        return jsonify({"error": str(e)}), 500
+
+if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=8000, debug=True)  # Listen on all interfaces