Update app.py

app.py

@@ -1,9 +1,20 @@
 import time
 import gradio as gr
+from ctransformers import AutoModelForCausalLM
 
-def slow_echo(message, history):
-    for i in range(len(message)):
-        time.sleep(0.3)
-        yield "You typed: " + message[: i+1]
+# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
+model_path = "WizardLM-7B-uncensored.Q3_K_S.gguf"
+llm = AutoModelForCausalLM.from_pretrained(model_path, model_type="llama")
 
-gr.ChatInterface(slow_echo).launch()
+def generate_response(message):
+    response = llm(message)
+    yield response
+
+def chatbot(message, history):
+    response_generator = generate_response(message)
+    for response in response_generator:
+        time.sleep(0.3)  # Optional delay for a natural chat feel
+        yield response
+
+iface = gr.ChatInterface(chatbot)
+iface.launch()