Skip to content

LINGO LABSIIT Gandhinagar

Models

Ansh

Multilingual Indic tokenizers with 128k, 160k, and 256k vocabularies.

Theme: Foundation Model Architectures & Efficient Intelligence
Type: Tokenizers
Funding: Unfunded

Releases

Ansh-128k, 160k, 256k

Open tokenizers on Hugging Face.

128k 160k 256k

Theme →Datasets and models →Publications →