Thank you for your solid work. The gso_1M dataset is close to 700G, which is a huge amount of data. How large is the subset (gso_1M) used by GSPose? If possible, can you provide the download link for the subset directly?
How long does it take to train using 2 batches on RTX3090?
Thank you very much.