Experimental Setup

[YunzhuoSun]

Dear Yan Yang, Liyuan Pan, and Liu Liu,

I recently read your paper titled "Event Camera Data Pre-training" and found it to be quite insightful. Firstly, I'd like to commend you on your excellent work in this area.

While going through your paper, I had a question regarding the experimental setup. Could you specify how many GPU resources were utilized for the pre-training process? This information would greatly assist me in understanding and replicating your results.

Thank you in advance for your time and consideration. I look forward to hearing back from you.

Best regards,
Yunzhuo Sun

[Yan98]

Hi Yunzhuo Sun,

Thank you for your interest. It is dependent on the masking ratio. With a high masking ratio, you should able to train it with 8 GPUs that have memory bigger than 16GB.

Hope this will help.