-
Hi, very helpful toolkit, I have learned a lot from it.
Recently, I have been focused on the multi-lingual title generation related tasks, and found that xProphetNet model has good performance, esp…
-
Hi again and thanks for your help earlier on huggingface/datasets#4 .
I managed to get the data downloaded and some models trained. However, I'm a bit unclear about what settings were used to comp…
-
### Ticket Contents
Belongg is developing BelonggAI, a tool that will help development practitioners, researchers, funders, etc analyze their proposals, program documents, policy documents, etc to …
-
### I want to finetune existing M2M 100 models (418M, 1.2B, 12B) for a particular language pair translation say (bn-en).
Which script(s) or training sequence/tutorial/doc should I follow. Currently…
-
In this issue you can either:
- Add papers that you think are interesting to read and discuss (please stick to the format).
- vote: should be done using :+1: on comments
Example: https://githu…
-
## 🐛 Bug
The [collater function](https://github.com/pytorch/fairseq/blob/master/fairseq/data/denoising_dataset.py#L355) for `DenoisingDataset` does not set the correct EOS token when used in the Mu…
-
Hi,
I would like to use a specific transformer encoder model such as `roberta-large` instead of the default `roberta`, which loads a `roberta-base`. Is there any way to do so?
-
Hi, When I use the `train_mbart_model.sh` to get further pre train based on the mBart-large-50 from this:https://huggingface.co/facebook/mbart-large-50;
And when I run in single GPU, there is no p…
-
Dependent on info from :
- existing code base
- newest text version
- log of past decisions
~Outcome - Answers to the following questions:~
~- [ ] do we need to revise any previous decisions about …
-
Currently, some predicted relations are unexpected or, instead, some expected relations are missed in answers of the system.
We need to detect the examples and try to solve them.