-
Hello, can you share with me the wmt19 model you distilled and trained?
-
Hi, I have been reproducing your results on IWSLT-16 En-De experiments using the NAT pre-trained models. However, I get different result when I use different batch_size.
- When batch_size = 1:
!…
-
You have to use the most recent .npy schedule file saved before .pt model weight file.
For this sentence, it means Must the pt weight file be greater than the npy noise schedule file? Can't you use t…
-
你好,原先的数据集下载链接好像找不着数据集了(404),请问可以更新一下数据集链接吗~~
-
Hello! Thanks for your code sharing.
When I run your default inference code:
bash ./inference_scrpts/iwslt_inf.sh path-to-ckpts/ema_0.9999_280000.pt path-to-save-results path-to-ckpts/alpha_cumpro…
-
Hi
Thanks for this great work, but when I reproduce the training processing followed by Readme, the arch cannot find in the fairseq. Can you kindly help with this?
![image](https://user-images.gi…
-
Can you give some comparative experiment results between this one and Tensorflow one? Do these two performs similar?
-
For pretrained models, BLEU for "WMT" and "TED" are reported:
|Language Pair | WMT | TED |
|--- |--- |--- |
|EN-DE |42.1 | 33.0
|EN-ES |35.9 | 37.0
|EN-HI | 22.3 | -
|EN-IT |31.8 | 31.7
|EN…
-
Not sure what is going on here but the best that I can tell is that there is a gzip file that seems to be missing.
Thank You
Tom
Traceback (most recent call last):
File "/home/tom/anacon…
-
## 🐛 Bug
**Describe the bug**
I came across this error when using **data.Field**. It only happen when I define my own **unk_token** and set **min_freq** >1 at the same time.
**To Reproduce**
the…