Update app.py

app.py

@@ -1,18 +1,22 @@
 import gradio as gr
-from optimum.pipelines import pipeline
-from transformers import AutoTokenizer
+from sentence_transformers import CrossEncoder
+from fastrag.utils import optimize_model
 
-# Load ONNX optimized model
-model_name = "jinaai/jina-reranker-v2-base-multilingual"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = pipeline("text-classification", model=model_name, tokenizer=tokenizer, trust_remote_code=True)
+# Load and optimize the model
+model = CrossEncoder(
+    "jinaai/jina-reranker-v2-base-multilingual", 
+    trust_remote_code=True
+)
+
+# Apply Intel IPEX optimization (FastRAG)
+model = optimize_model(model, backend="ipex")  # ✅ Faster CPU inference
 
 # Function to rerank documents
 def rerank(query, documents):
-    documents = documents.split("&&&")
+    documents = documents.split("&&&")  # Use special delimiter
     inputs = [[query, doc] for doc in documents if doc.strip()]
-    scores = model(inputs)
-    ranked_docs = sorted(zip(documents, [s['score'] for s in scores]), key=lambda x: x[1], reverse=True)
+    scores = model.predict(inputs)
+    ranked_docs = sorted(zip(documents, scores), key=lambda x: x[1], reverse=True)
     return [{"document": doc, "score": round(score, 4)} for doc, score in ranked_docs]
 
 # Gradio Interface