search

duyichao / E2E-ST-TDA

Official implementation of AAAI'2022 paper "Regularizing End-to-End Speech Translation with Triangular Decomposition Agreement"
MIT License
18 stars 3 forks source link

Vocabulary is unavailable in the released well-trained models #3

Open bzhangGo opened 2 years ago

bzhangGo commented 2 years ago

Hi all,

Thanks for this nice work and for releasing the trained models!

I'd like to do some experimental analysis and compare my work with your models. Could you please also release the TDA-m model as well as the vocabulary files to let others reproduce your model output? Or, alternatively, could you please offer the model's output translation directly (this might be much easier, and I could compare the translation of my approach with your translation readily)?

Thanks,

duyichao commented 2 years ago

Hi Biao,

We are very sorry that the checkpoints and vocabulary files were not uploaded in time because the lab server storage was down. The good news is that our server storage will be repaired soon (this week). If it goes well, I will update all checkpoints and vocabulary files in the next little while. 16611649948597_ pic_hd

Best regards,