Closed zhengyuan-xie closed 2 months ago
We mainly conduct all experiments using 4 V100 GPUs. We also checked that the performance difference between using 4 V100s and 8 V100s was minimal.
We mainly conduct all experiments using 4 V100 GPUs. We also checked that the performance difference between using 4 V100s and 8 V100s was minimal.
Thanks for your fast reply!
Hi, thanks for your work! I want to reproduce the result and wonder the number of GPUs used in these experiments and the training time, hope you can help me!