This is a smaller version of the google/mt5-large model with only Hungarian and some English embeddings left.

After shrinking the sentencepiece vocabulary from 250K to 50K (top 10K English and top 40K Hungarian tokens) the model size reduced from 4.92GB to 3.28GB.

This model is based on the description of cointegrated/rut5-base. The creation of this model is described in the post How to adapt a multilingual T5 model for a single language along with the source code.

Downloads last month: 18

Safetensors

Model size

820M params

Tensor type

F32

Inference Examples

Text2Text Generation

This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.