Make request async and show that request is being processed

README.md

@@ -16,7 +16,7 @@ The clients depend on the following Python packages:
 To install the dependencies, run the following command:
 
 ```bash
-pip install openai tkhtmlview
+pip install openai tkhtmlview tqdm
 ```
 
 ## Usage

cli.py

@@ -1,8 +1,8 @@
 #!/usr/bin/env python3
 import os
-import readline
-
+import asyncio
 import openai
+from tqdm import tqdm
 
 MODEL = "gpt-4"
 
@@ -22,7 +22,29 @@
 def user_input_dict(prompt):
  return {"role": "user", "content": prompt}
 
+
+async def get_response_async():
+ loop = asyncio.get_event_loop()
+ response_future = loop.run_in_executor(None, lambda: openai.ChatCompletion.create(
+ model=MODEL, messages=messages, temperature=0.1))
+
+ # Display dotting
+ pbar = tqdm(total=1, bar_format='{desc}', position=0, leave=True)
+ x = 0
+ while not response_future.done():
+ pbar.set_description_str("." * x, refresh=True)
+ x += 1
+ await asyncio.sleep(0.5)
+
+ pbar.close()
+
+ return await response_future
+
+
 price = 0
+
+loop = asyncio.get_event_loop()
+
 while True:
  user_input = input("You: ")
  if not user_input:
@@ -32,8 +54,7 @@ def user_input_dict(prompt):
 
  messages.append(user_input_dict(user_input))
 
- response = openai.ChatCompletion.create(
- model=MODEL, messages=messages, temperature=0.1)
+ response = loop.run_until_complete(get_response_async())
 
  message = response.choices[0]["message"]
  messages.append(message)