Update app.py

app.py

@@ -7,12 +7,23 @@ import numpy as np
 import torch
 from sklearn.neighbors import NearestNeighbors
 from sentence_transformers import SentenceTransformer
+from transformers import pipeline, AutoTokenizer, AutoModelForCausalLM
 
 # --- CONFIGURATION ---
-HF_TOKEN = os.getenv("HF_TOKEN", "").strip()
-HF_MODEL = "HuggingFaceH4/zephyr-7b-beta"  # Change if you want
-HF_API_URL = f"https://api-inference.huggingface.co/models/{HF_MODEL}"
-headers = {"Authorization": f"Bearer {HF_TOKEN}"}
+#HF_TOKEN = os.getenv("HF_TOKEN", "").strip()
+#HF_MODEL = "HuggingFaceH4/zephyr-7b-beta"  # Change if you want
+#HF_API_URL = f"https://api-inference.huggingface.co/models/{HF_MODEL}"
+#headers = {"Authorization": f"Bearer {HF_TOKEN}"}
+
+
+
+print("🔄 Loading local Falcon model...")
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+falcon_model = AutoModelForCausalLM.from_pretrained(model_name)
+pipe = pipeline("text-generation", model=falcon_model, tokenizer=tokenizer)
+
+
+
 
 FILES = ["main1.txt", "main2.txt", "main3.txt", "main4.txt", "main5.txt", "main6.txt"]  # Your text files
 EMBEDDING_MODEL = "sentence-transformers/all-MiniLM-L6-v2"  # Light and fast
@@ -90,33 +101,21 @@ User Question: {question}
 Answer:"""
     return prompt
 
-def respond(message, history):
-    try:
-        prompt = build_prompt(message)
-
-        payload = {
-            "inputs": prompt,
-            "parameters": {"temperature": 0.2, "max_new_tokens": 400},
-        }
-
-        response = requests.post(HF_API_URL, headers=headers, json=payload, timeout=30)
-        response.raise_for_status()
-        output = response.json()
-        generated_text = output[0]["generated_text"]
-        match = re.search(r"Answer:(.*)", generated_text, re.DOTALL)
-        answer = generated_text[len(prompt):].strip()
-
 
 
+def respond(message, history):
+    prompt = build_prompt(message)
+    try:
+        output = pipe(prompt, max_new_tokens=300, temperature=0.2)
+        answer = output[0]["generated_text"].split("Answer:")[-1].strip()
     except Exception as e:
-        print("API Error:", e)
-        answer = "❌ Error contacting the model. Please try again later."
+        print("Error:", e)
+        answer = "❌ Error generating response."
 
     if history is None:
         history = []
-
-    history.append({"role": "assistant", "content": answer})
-    return answer
+    history.append((message, answer))
+    return answer, history
 
 # --- INIT SECTION ---