Upload 2 files

app.py

@@ -5,8 +5,8 @@ import numpy as np
 from typing import List, Optional
 from langchain_community.vectorstores import FAISS
 from langchain.embeddings.base import Embeddings
-from gradio_client import Client
 import gradio as gr
+from sentence_transformers import SentenceTransformer
 
 # Configuration
 DATA_FILE = "data-mtc.txt"  # This file is no longer used in the Space
@@ -38,30 +38,17 @@ logging.basicConfig(
     ]
 )
 
-# Example Questions Pool
-EXAMPLE_QUESTIONS = [
-    "Comment intégrer les enseignements MTC dans la vie quotidienne ?",
-    "Comment se préparer à une discussion de groupe MTC ?",
-    "Quels sont les obstacles courants à la compréhension des Chroniques ?"
-]
+# Embedding Model Integration
+device = torch.device("cpu")
+embedding_model = SentenceTransformer("Snowflake/snowflake-arctic-embed-l", device=device, trust_remote_code=True)
 
-class GradioEmbeddings(Embeddings):
-    """Embedding management using Gradio API"""
-
-    def __init__(self):
-        super().__init__()
-        self.client = Client("localsavageai/embijiji3")
+class HuggingFaceEmbeddings(Embeddings):
+    """Embedding management using Hugging Face SentenceTransformer"""
 
     def _generate_embedding(self, text: str) -> np.ndarray:
-        """Generate an embedding via the Gradio API"""
+        """Generate an embedding via the Hugging Face model"""
         try:
-            result = self.client.predict(
-                document=text.strip(),
-                api_name="/embed"
-            )
-            if not isinstance(result, list):
-                raise ValueError("Invalid embedding response from Gradio API")
-            return np.array(result, dtype=np.float32)
+            return np.array(embedding_model.encode(text.strip()), dtype=np.float32)
         except Exception as e:
             logging.error(f"Embedding error: {str(e)}")
             raise RuntimeError("Failed to generate embedding") from e
@@ -75,7 +62,7 @@ class GradioEmbeddings(Embeddings):
 
 def initialize_vector_store() -> FAISS:
     """Robust initialization of the vector store"""
-    embeddings = GradioEmbeddings()
+    embeddings = HuggingFaceEmbeddings()
 
     try:
         logging.info("Loading existing database...")
@@ -114,21 +101,28 @@ def generate_response(user_input: str, vector_store: FAISS) -> Optional[str]:
             for i, doc in enumerate(best_docs)
         )
 
-        response = Client("Qwen/Qwen2.5-Max-Demo").predict(
-            query=user_input,
-            history=[],
-            system=BASE_SYSTEM_PROMPT.format(context=context),
-            api_name="/model_chat"
-        )
+        from transformers import AutoModelForCausalLM, AutoTokenizer
 
-        if isinstance(response, tuple) and len(response) >= 2:
-            chat_history = response[1]
-            if isinstance(chat_history, list) and len(chat_history) > 0:
-                last_message = chat_history[-1]
-                if isinstance(last_message, (list, tuple)) and len(last_message) >= 2:
-                    return last_message[1]
+        model_name = "Qwen/Qwen2.5-72B-Instruct"
+        model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
+        tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+        prompt = BASE_SYSTEM_PROMPT.format(context=context)
+        
+        messages = [
+            {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
+            {"role": "user", "content": user_input}
+        ]
 
-        return "Réponse indisponible - Veuillez reformuler votre question."
+        text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+        
+        model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
+        
+        generated_ids = model.generate(**model_inputs, max_new_tokens=512)
+        
+        response = tokenizer.batch_decode(generated_ids[:, model_inputs.input_ids.shape[-1]:], skip_special_tokens=True)
+        
+        return response[0] if response else "Réponse indisponible - Veuillez reformuler votre question."
 
     except Exception as e:
         logging.error(f"Generation error: {str(e)}", exc_info=True)
@@ -146,66 +140,18 @@ def chatbot(query):
         return f"Une erreur s'est produite : {str(e)}"
 
 
-# Rotating Example Questions Functionality
-def get_random_questions():
-    """Selects three random example questions"""
-    return random.sample(EXAMPLE_QUESTIONS, 3)
-
 # Gradio Interface Setup with Enhanced UI
 with gr.Blocks(title="MTC Chatbot") as demo:
     gr.Markdown("# Apprenez-en plus sur le savoir MTC!")
     
     chatbot_ui = gr.Chatbot(label="MTC Assistant", type="messages")
     
-    with gr.Row():
-        input_box = gr.Textbox(
-            placeholder="Posez votre question ici...",
-            label="Votre question"
-        )
+    input_box = gr.Textbox(
+        placeholder="Posez votre question ici...",
+        label="Votre question"
+    )
     
-    def respond(message, history):
-        vs = initialize_vector_store()
-        response = generate_response(message, vs)
-        
-        history.append({"role": "user", "content": message})
-        history.append({"role": "assistant", "content": response})
-        
-        # After every interaction, get new random questions
-        example_questions = get_random_questions()
-        
-        # Recreate the buttons with new questions
-        example_buttons = []
-        for question in example_questions:
-            btn = gr.Button(question)
-            btn.click(
-                process_example_click,
-                inputs=[gr.Textbox(value=question, visible=False), chatbot_ui],
-                outputs=chatbot_ui
-            )
-            example_buttons.append(btn)
-        
-        return history
-
-    def process_example_click(example_query, history):
-        response = chatbot(example_query)
-        history.append({"role": "user", "content": example_query})
-        history.append({"role": "assistant", "content": response})
-        return history
-
-    # Initial example questions
-    example_questions = get_random_questions()
-    with gr.Row():
-        example_buttons = []
-        for question in example_questions:
-            btn = gr.Button(question)
-            btn.click(
-                process_example_click,
-                inputs=[gr.Textbox(value=question, visible=False), chatbot_ui],
-                outputs=chatbot_ui
-            )
-            example_buttons.append(btn)
-
-    input_box.submit(respond, [input_box, chatbot_ui], chatbot_ui)
+    input_box.submit(chatbot, inputs=input_box, outputs=chatbot_ui)
 
 if __name__ == "__main__":
     demo.launch()

requirements.txt

@@ -4,4 +4,6 @@ faiss-cpu
 gradio
 gradio_client
 numpy
-
+sentence_transformers
+einops
+torch