Hi, nice work! It looks like the dataset is very large, with 1.2TB of raw data and 41TB of processed data. Do I need to use all the data if I want to get the results in the paper? Moreover, which GPUs are you using for training? Is it a single-GPU training? What is the approximate duration of the training? Thank you very much.
Hi, nice work! It looks like the dataset is very large, with 1.2TB of raw data and 41TB of processed data. Do I need to use all the data if I want to get the results in the paper? Moreover, which GPUs are you using for training? Is it a single-GPU training? What is the approximate duration of the training? Thank you very much.