We improved on the following features. * Text line-break within <pre> * Allowed tokenizer settings with huggingface tokenizer, from langhuan.loaders * Preset tags before tagging
And we fixed bugs * Allow manually set index, before there will be bug on langhuan.options.Options.df
0.0.8
Features 🚀 * Changed format on downloaded data * Downloaded NER data can be transformed to a **pytorch dataset/ dataloader** with a certain [HuggingFace tokenizer](https://github.com/huggingface/tokenizers), see the detailed [documentation](https://raynardj.github.io/langhuan/docs/loader/) Fix 🐛 * Offset error caused by ◀️ and ▶️
0.0.6
1st Stable Version * Support for NERTask, ClassifyTask * Setup data from Pandas DataFrame * Allow multi-user verification * Saving to & loading from history * Ordering strategies * march forward (default) * pincer * trident * Admin page to monitor progress and download data, manage options * Tested frontend on late versions of Chrome, FireFox & Safari