XLM-RoBERTa is not supported

[hammad26]

Hi Thanks a lot for your hard work. Can you add support XLM-RoBERTa too? Thanks

[kinoute]

Any news from this one? It seem really good but maybe it needs a lot of work to add it here.

Flair is about to add it. Here's the PR, you can see some code that could help you to add it: https://github.com/flairNLP/flair/pull/1333

[ThilinaRajapakse]

Shouldn't be too hard I think. I should be able to add this soon.

[ThilinaRajapakse]

Added in 0.17.0. Let me know if you run into any issues! :wink:

[kinoute]

Tested the model quickly this morning. Seems good so far.

[alejandrojcastaneira]

Hi. I'm initializing the model class like this;

model = MultiLabelClassificationModel('xlm-roberta', 'xlm-roberta-base')

however I get this error:

line 69, in init config_class, model_class, tokenizer_class = MODEL_CLASSES[model_type] KeyError: 'xlm-roberta'

Is the model also supported for multiple label classification?

[kinoute]

Try xlmroberta.

[alejandrojcastaneira]

I tried, it gives the same error: KeyError: 'xlmroberta'

[ThilinaRajapakse]

It's not added for multi-label classification at the moment.

[alejandrojcastaneira]

hello @ThilinaRajapakse just to be sure, it's XLM_Roberta already available in the model's list for multiple label classification? We had good results with other models and we are looking forward to try this one, we are very excited!

[ThilinaRajapakse]

It's just been added in 0.22.1.

[alejandrojcastaneira]

thanks you @ThilinaRajapakse so much, for your support!

[freeIsa]

Is there a reason why xml-roberta is not supported for language modeling fine-tuning?