def get_chat_response()

in lmms_eval/tasks/mmvet/utils.py [0:0]

• 35 lines of code
• 5 McCabe index (conditional complexity)

def get_chat_response(prompt, model=GPT_EVAL_MODEL_NAME, temperature=0.0, max_tokens=128, patience=3, sleep_time=5):
    headers = {
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json",
    }

    messages = [
        {"role": "user", "content": prompt},
    ]

    payload = {
        "model": model,
        "messages": messages,
        "temperature": temperature,
        "max_tokens": max_tokens,
    }

    while patience > 0:
        patience -= 1
        try:
            response = requests.post(
                API_URL,
                headers=headers,
                json=payload,
                timeout=60,
            )
            response.raise_for_status()
            response_data = response.json()

            content = response_data["choices"][0]["message"]["content"].strip()
            if content != "":
                return content, response_data["model"]

        except Exception as e:
            eval_logger.error(f"Error: {e}")
            if "Rate limit" in str(e):
                eval_logger.info("Sleeping due to rate limit...")
                time.sleep(sleep_time)
            eval_logger.info(f"Retrying...Patience left: {patience}")

    return "", ""