metadata
language: ti
widget:
- text: ዓቕሚ ደቀንስትዮ [MASK] ብግብሪ ተራእዩ
BERT Base for Tigrinya Language
We pretrain a BERT base-uncased model on a relatively small dataset for Tigrinya (34M tokens) for 40 epochs.
Contained in this card is a PyTorch model exported from the original model that was trained on a TPU v3.8 with Flax.
Hyperparameters
The hyperparameters corresponding to model sizes mentioned above are as follows:
| Model Size | L | AH | HS | FFN | P |
|---|---|---|---|---|---|
| BASE | 12 | 12 | 768 | 3072 | 110M |
(L = number of layers; AH = number of attention heads; HS = hidden size; FFN = feedforward network dimension; P = number of parameters.)