12-layers, 768-hidden, 12-heads, 110M parameters:
- lvbert-base-tf (1.1GB)
- lvbert-base-pytorch (394MB)
Znotiņš, A., Bārzdiņš, G. LVBERT: Transformer-based Model for Latvian Language Understanding. Human Language Technologies - The Baltic Perspective. Frontiers in Artificial Intelligence and Applications, vol. 328, IOS Press, 2020
- NLP-PIPE: Latvian NLP Pipeline as a Service
- Full Stack of Latvian Language Resources for NLU and NLG
This work is supported by the Latvian Council of Science under the grant agreement No. lzp-2018/2-0216 (Latvian Language Understanding and Generation in Human-Computer Interaction) and by the State Research Programme under the grant agreement No. VPP-IZM-DH-2020/1-0001 (Digital Resources for Humanities: Integration and Development).
LVBERT by AiLab is licensed under a GNU General Public License v3.0.
Please contact us if you are interested in acquiring a different licence.
Please, cite the relevant publications if you use this software or models in your research. Please, let us know if you use this software or models in your products or services. Your citations and feedback are important to secure funding for the further development of LVBERT.