I've download the dunuhuang pre-trained model from your drive, and I saw your iteration is about 300k. But when I tried to load the checkpoints and continue training on your work, I found that the iteration start from 24000, is that correct?
And what should I set for the 'MASK' config in config.yml if I try to train on dunhuang dataset?
By the way, why my testing result using your pre-trained model on dunhuang dataset is not good as you show in the paper?
Thank you for your work.
Looking forward to your reply.
I've download the dunuhuang pre-trained model from your drive, and I saw your iteration is about 300k. But when I tried to load the checkpoints and continue training on your work, I found that the iteration start from 24000, is that correct?
And what should I set for the 'MASK' config in config.yml if I try to train on dunhuang dataset?
By the way, why my testing result using your pre-trained model on dunhuang dataset is not good as you show in the paper?
Thank you for your work.
Looking forward to your reply.