.PT

isolezwe_isizulu_AWD_wt2InspiredStuBiggerP100.pt (227.55 MB)

.PT

isolezwe_isizulu_basicInputDrop.pt (25.12 MB)

.PT

nchlt_isizulu_AWD_wtInspiredStuBiggerP100.pt (227.55 MB)

.PT

nchlt_isizulu_basicInputDrop.pt (25.12 MB)

.PT

nchlt_sepedi_AWD_wtInspiredStuBiggerP100.pt (227.55 MB)

.PT

nchlt_sepedi_basicInputDrop.pt (25.12 MB)

.PT

nchlt_isizulu_QRNN_wt2InspiredQRNNLongTrain.pt (256.61 MB)

.PT

nchlt_sepedi_QRNN_wtInspiredStuBiggerP100mk2.pt (227.31 MB)

.PT

isolezwe_isizulu_QRNN_wtInspiredStuBiggerP100mk2.pt (126.15 MB)

Pytorch Language Models trained on South African Languages

Version 3 2021-04-06, 07:55

Version 2 2021-04-01, 14:21

Version 1 2021-03-31, 12:54

software

posted on 2021-04-06, 07:55 authored by Luc HaywardLuc Hayward, Jan Buys, Stuart MeshamStuart Mesham

Pytorch Language Models trained on the South African isiZulu and Sepedi languages using the NCHLT and Isolezwe datasets.

AWD-LSTM, QRN, Basic LSTM and Transformer architectures are each included for each of the datasets used.

History

University of Cape Town Computer Science Department

RefWorks

BibTeX

Ref. manager

Endnote

DataCite

NLM