Intended Use Cases
token-classification
Restricted — review the source terms before use.
Every column the registry tracks. Values marked derivedwere filled by a rule or read from the source's text rather than stated in its metadata; the rule or the quoted sentence is shown under the value.
This is the same multilingual corpus, sliced for other languages the platform indexes it under.
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
NER model · XLMRobertaForTokenClassification · Jesujoba Alabi, David Adelani et al. (Masakhane)
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
Shona (sna) achieves F1=96.5 on MasakhaNER 2.0 — highest in the benchmark. Multilingual NER model covering 21 African languages.
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
token-classification model · XLMRobertaForTokenClassification · masakhane
XLM-RoBERTa fine-tuned on Nigerian Pidgin (pcm) corpus. Outperforms multilingual XLM-R on NER tasks for Nigerian Pidgin.
mBERT fine-tuned on Nigerian Pidgin (pcm). Better NER performance than multilingual baseline.
Multilingual sentiment model fine-tuned on NaijaSenti for Nigerian Pidgin and 3 other Nigerian languages. SOTA on NaijaSenti benchmark at time of publication.