Fubang ZHAO

no lr_layer_decay_rate for embedding

3

Thanks for your work. I found that there is no lr_layer_decay_rate for embedding layer, which is weird because embedding is actually below transformer layers.

why not use a partial factorization ?

7

Hi, thanks for the great work. I have a question about the process of pre-training. Assuming a lenth-T sequence with a permutation z, since we use a partial prediction(only predict...

Hi，校友们～ 我最近在尝试用你们提供的mel数据做实验，非常好的工作！ 想请问，这几个数据集你们有baseline的方案和F1可以report吗？我想知道自己做到什么程度了。 Thanks

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction  ### Expected behavior _No response_ ### System Info _No response_ ### Others _No...

您好，非常好的工作，但是我有一个小问题想请教一下。 在论文的4.2节，您写到： “By thorough error analysis, we realize that for the order h-t-r (t-h-r follows the same logic), the model has to predict all t with regard to h in...

check了一下，发现有40%都是重复的，是不是数据上传有什么问题？