Function get_eval

light-eval/src/eval_llavabenchmark.py:178–200

(content: str, max_tokens: int)

    

def get_eval(content: str, max_tokens: int):
    while True:
        try:
            response = openai.ChatCompletion.create(
                model='gpt-4-0314',
                messages=[{
                    'role': 'system',
                    'content': 'You are a helpful and precise assistant for checking the quality of the answer.'
                }, {
                    'role': 'user',
                    'content': content,
                }],
                temperature=0.2,  # TODO: figure out which temperature is best for evaluation
                max_tokens=max_tokens,
            )
            break
        except openai.error.RateLimitError:
            pass
        except Exception as e:
            print(e)
        time.sleep(NUM_SECONDS_TO_SLEEP)

    return response['choices'][0]['message']['content']


def parse_score(review):