XLM-RoBERTa Base for Luganda (xlm_roberta_base_finetuned_luganda)

Description

xlm_roberta_base_finetuned_luganda is a ** Luganda RoBERTa** model obtained by fine-tuning xlm-roberta-base model on Luganda language texts. It provides better performance than the XLM-RoBERTa on named entity recognition datasets.

Specifically, this model is an xlm-roberta-base model that was fine-tuned on the Luganda corpus.

Predicted Entities

Download

How to use

embeddings = XlmRoBertaEmbeddings.pretrained("xlm_roberta_base_finetuned_luganda", "lg") \
      .setInputCols("sentence", "token") \
      .setOutputCol("embeddings")
val embeddings = XlmRoBertaEmbeddings.pretrained("xlm_roberta_base_finetuned_luganda", "lg")
      .setInputCols("sentence", "token")
      .setOutputCol("embeddings")

Model Information

Model Name: xlm_roberta_base_finetuned_luganda
Compatibility: Spark NLP 3.3.0+
License: Open Source
Edition: Official
Input Labels: [token, sentence]
Output Labels: [embeddings]
Language: lg
Case sensitive: true

Data Source

https://huggingface.co/Davlan/xlm-roberta-base-finetuned-luganda

Benchmarking

## Eval results on Test set (F-score, average over 5 runs)

Dataset| XLM-R F1 | lg_roberta F1
-|-|-
[MasakhaNER](https://github.com/masakhane-io/masakhane-ner) | 79.69 | 84.70