Benedikt Fuchs
To be removed, once it is done: Please add the appropriate label to this ticket, e.g. feature or enhancement. **Is your feature/enhancement request related to a problem? Please describe.** ONNX...
First PR about https://github.com/flairNLP/flair/issues/2640: refactoring models, such that each model has a `_prepare_tensors` and a `forward` method, where one extracts all tensors out of the data points and the latter...
since some refactoring, NER labels will be added as span labels instead of following the BIO or BIOES tagging shemas. This PR updates the JsonlCorpus to follow the newer way...
adds ACE, from this repo: https://github.com/Alibaba-NLP/ACE currently not tested on larger training, but at least it runs through when using CONLL_03 downsampled to 0.005 on each dataset. ## Inner working:...
A proposal to return the loss per sentence. This way, it would be easier to compute uncertainty scores used for [active learning](https://arxiv.org/abs/1707.05928). I made experience in prior projects, that that...
Hello and thank you for sharing this interesting approach. I have one question regarding dropout. If I understand the published code correctly, MeZO was tested having dropout deactivated, e.g. lines...
closes https://github.com/flairNLP/flair/issues/3403
Closes gh-3243 A suggestion how we could support pickle and deepcopy without having to create multiple spans etc. The idea is to use `__getitem__` to create spans and relations by...