Set pad_token to eos_token and exclude user query from response

app.py

@@ -1,10 +1,11 @@
 import gradio as gr
 import requests
 from typing import List, Dict, Tuple
-from flask import Flask, request, jsonify
+from flask import Flask, request, jsonify, send_from_directory
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import threading
 import torch
+import os
 
 # Define the API URL to use the internal server
 API_URL = "http://localhost:5000/chat"
@@ -23,6 +24,11 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 if tokenizer.pad_token is None:
     tokenizer.pad_token = tokenizer.eos_token
 
+# Route to serve static files (e.g., images)
+@app.route('/static/<path:path>')
+def send_static(path):
+    return send_from_directory('static', path)
+
 @app.route('/chat', methods=['POST'])
 def chat():
     data = request.json
@@ -138,8 +144,8 @@ with gr.Blocks(css='''
 ''', js=None) as demo:
     gr.Markdown("""
 <div class="dicta-header">
-  <a href="">
-    <img src="\\logo111.png" alt="Logo" class="dicta-logo">
+  <a href="/static/logo111.png">
+    <img src="/static/logo111.png" alt="Logo" class="dicta-logo">
   </a>  
   <div class="dicta-intro-text">
     <h1>צ'אט מערכי - הדגמה ראשונית</h1>