Merge pull request #73 from openai/dev/logan/add-assistants

Add assistants example and move to the multi example structure

Author: logankilpatrick

README.md

@@ -1,6 +1,6 @@
-# OpenAI API Quickstart - Python example app
+# OpenAI API Quickstart - Python
 
-This is an example chat app intended to get you started with your first OpenAI API project. It uses the [Chat Completions API](https://platform.openai.com/docs/api-reference/chat) to create a simple general purpose chat app with streaming.
+This repository hosts multiple quickstart apps for different OpenAI API endpoints (chat, assistants, etc). Check out the `examples` folder to try out different examples and get started using the OpenAI API.
 
 ## Basic request
 
@@ -21,8 +21,6 @@ completion = client.chat.completions.create(
 print(completion.choices[0].message)
 ```
 
-This quickstart app builds on top of the example code above, with streaming and a UI to visualize messages.
-
 ## Setup
 
 1. If you don’t have Python installed, install it [from Python.org](https://www.python.org/downloads/).
@@ -38,6 +36,7 @@ This quickstart app builds on top of the example code above, with streaming and
 4. Create a new virtual environment:
 
    - macOS:
+
      ```bash
      $ python -m venv venv
      $ . venv/bin/activate
@@ -65,8 +64,16 @@ This quickstart app builds on top of the example code above, with streaming and
 
 8. Run the app:
 
-   ```bash
-   $ flask run
-   ```
+This step depends on the app itself. If the code uses flask (like the chat-basic example), you can run:
+
+```bash
+$ flask run
+```
 
 You should now be able to access the app from your browser at the following URL: [http://localhost:5000](http://localhost:5000)!
+
+If the code is just a simple Python script, you can run it with:
+
+```bash
+$ python my_file.py
+```

examples/assistant-basic/assistant.py

@@ -0,0 +1,79 @@
+from openai import OpenAI
+from time import sleep
+
+client = OpenAI()
+starting_assistant = ""
+starting_thread = ""
+
+
+def create_assistant():
+    if starting_assistant == "":
+        my_assistant = client.beta.assistants.create(
+            instructions="You are a helpful assistant.",
+            name="MyQuickstartAssistant",
+            model="gpt-3.5-turbo",
+        )
+    else:
+        my_assistant = client.beta.assistants.retrieve(starting_assistant)
+
+    return my_assistant
+
+
+def create_thread():
+    if starting_thread == "":
+        thread = client.beta.threads.create()
+    else:
+        thread = client.beta.threads.retrieve(starting_thread)
+
+    return thread
+
+
+def send_message(thread_id, message):
+    thread_message = client.beta.threads.messages.create(
+        thread_id,
+        role="user",
+        content=message,
+    )
+    return thread_message
+
+
+def run_assistant(thread_id, assistant_id):
+    run = client.beta.threads.runs.create(
+        thread_id=thread_id, assistant_id=assistant_id
+    )
+    return run
+
+
+def get_newest_message(thread_id):
+    thread_messages = client.beta.threads.messages.list(thread_id)
+    return thread_messages.data[0]
+
+
+def get_run_status(thread_id, run_id):
+    run = client.beta.threads.runs.retrieve(thread_id=thread_id, run_id=run_id)
+    return run.status
+
+
+def main():
+    my_assistant = create_assistant()
+    my_thread = create_thread()
+
+    while True:
+        user_message = input("Enter your message: ")
+        if user_message.lower() == "exit":
+            break
+
+        send_message(my_thread.id, user_message)
+        run = run_assistant(my_thread.id, my_assistant.id)
+        while run.status != "completed":
+            run.status = get_run_status(my_thread.id, run.id)
+            sleep(1)
+            print("⏳", end="\r", flush=True)
+
+        sleep(0.5)
+        response = get_newest_message(my_thread.id)
+        print("Response:", response.content[0].text.value)
+
+
+if __name__ == "__main__":
+    main()