Update app.py

app.py

@@ -36,8 +36,8 @@ if prompt := st.chat_input(f"مرحبا انا سبيدي , كيف استطيع
     # Load model only when user submits a prompt
     try:
         # Load the tokenizer and model with caching in the specified directory
-        tokenizer = AutoTokenizer.from_pretrained("joermd/llma-speedy", cache_dir=cache_dir)
-        model = AutoModelForCausalLM.from_pretrained("joermd/llma-speedy", cache_dir=cache_dir)
+        tokenizer = AutoTokenizer.from_pretrained("sambanovasystems/SambaLingo-Arabic-Chat", cache_dir=cache_dir)
+        model = AutoModelForCausalLM.from_pretrained("sambanovasystems/SambaLingo-Arabic-Chat", cache_dir=cache_dir)
         
         # Generate response
         inputs = tokenizer(prompt, return_tensors="pt")