import httpx
import os
import asyncio
import logging
logging.basicConfig(level=logging.INFO)
from dotenv import load_dotenv
load_dotenv()
from starlette.responses import StreamingResponse
import json
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY") # recommended for security
async def run(prompt: str, max_retries: int = 3):
if not prompt:
return {"error": "No prompt provided"}
#logging.info(f"The value of my_variable is: {OPENAI_API_KEY}")
url = "https://api.openai.com/v1/chat/completions"
headers = {
"Authorization": f"Bearer {OPENAI_API_KEY}",
"Content-Type": "application/json"
}
payload = {
"model": "gpt-3.5-turbo", # or "gpt-4"
"messages": [
{"role": "system", "content": "You are a helpful yoga assistant."},
{"role": "user", "content": prompt}
],
"temperature": 0.7
}
retries = 0
while retries <= max_retries:
try:
async with httpx.AsyncClient(timeout=20.0) as client:
response = await client.post(url, headers=headers, json=payload)
if response.status_code == 429:
await asyncio.sleep(2 ** retries) # exponential backoff
retries += 1
continue
response.raise_for_status()
data = response.json()
return {
"response": data["choices"][0]["message"]["content"]
}
except httpx.HTTPError as e:
return {"error": str(e)}
return {"error": "Exceeded retry limit due to rate limiting"}
async def stream(prompt: str):
url = "https://api.openai.com/v1/chat/completions"
headers = {
"Authorization": f"Bearer {OPENAI_API_KEY}",
"Content-Type": "application/json"
}
payload = {
"model": "gpt-3.5-turbo",
"messages": [
{"role": "system", "content": "You are a helpful yoga assistant."},
{"role": "user", "content": prompt}
],
"stream": True,
"temperature": 0.7,
"max_tokens": 200
}
async def event_stream():
async with httpx.AsyncClient(timeout=60.0) as client:
async with client.stream("POST", url, headers=headers, json=payload) as response:
async for line in response.aiter_lines():
if line.startswith("data: "):
payload = line[len("data: "):]
if payload.strip() == "[DONE]":
break
try:
chunk = json.loads(payload)
token = chunk["choices"][0]["delta"].get("content", "")
yield token
except Exception:
continue
return StreamingResponse(event_stream(), media_type="text/plain")
