OpenAI API #102
Description
Hello, thank you for your outstanding work. In "eval_activitynet_qa.py", LLaMA-VID evaluates the predicted results by calling the OpenAI API. However, there are rate limits in my account when I call the API. So, I was wondering if there is any other method to evaluate the performance. Thanks!