Heming Xia

Results 16 comments of Heming Xia

If you encounter any problems, feel free to contact me 😊

Hi @ndkmath1, The training of the AT verifier is totally the same as the conventional Transformer in fairseq, just check the scripts below: 1. [fairseq/training-a-new-model](https://github.com/facebookresearch/fairseq/tree/main/examples/translation#training-a-new-model) 2. [scalingnmt/training-a-new-model-on-wmt16-en-de](https://github.com/facebookresearch/fairseq/blob/main/examples/scaling_nmt/README.md#training-a-new-model-on-wmt16-en-de)

Maybe [this issue](https://github.com/NVIDIA/NeMo/issues/2797) helps. Do you also have this problem when just running our code?

``` File "/home/user/.local/lib/python3.8/site-packages/omegaconf/omegaconf.py", line 669, in open_dict prev_state = config._get_node_flag("struct") AttributeError: 'dict' object has no attribute '_get_node_flag' ``` emmm it seems that the error comes from omegaconf. Could you try...

好的,期待您的进一步回复~🤝

> @hemingkx 请教一下 您可以复现LCM在 20NG这个数据集上的效果吗 基于bert的实验~ 没有,我们提出这个issue之后,就放弃这个任务了,欢迎讨论👏~