Update app.py

app.py

@@ -1,9 +1,27 @@
+# Install necessary libraries
+!pip install torch torchvision -f https://download.pytorch.org/whl/cu80/torch_stable.html
+!pip install gradio transformers
+!pip install tensorflow
+
+# Import required libraries
 import gradio as gr
 from transformers import pipeline
+import torch
+import threading
+import time
+import tensorflow as tf
+
+# Check GPU availability
+print(torch.cuda.is_available())
+print(tf.test.gpu_device_name())
 
 # Initialize the text generation pipeline with the specified model
 pipe = pipeline("text-generation", model="chargoddard/Yi-34B-Llama", device=0)
 
+# Rate limiting parameters
+rate_limit = 5  # Number of requests per second
+last_request_time = 0
+
 def respond(
     message,
     history: list[tuple[str, str]],
@@ -12,6 +30,15 @@ def respond(
     temperature,
     top_p,
 ):
+    global last_request_time
+
+    # Apply rate limiting
+    elapsed_time = time.time() - last_request_time
+    if elapsed_time < 1.0 / rate_limit:
+        time.sleep(1.0 / rate_limit - elapsed_time)
+
+    last_request_time = time.time()
+
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -53,5 +80,5 @@ demo = gr.ChatInterface(
     ],
 )
 
-if __name__ == "__main__":
-    demo.launch()
+# Launch the Gradio interface
+demo.launch()