This is a model derived from Twitter-roBERTa-base model trained on ~10K Discord messages from NFT-based Discord servers and finetuned for sentiment analysis with manually labelled data.
The original Twitter-roBERTa-base model can be found
here. This model is suitable for English.