File size: 314 Bytes
bd6529f
 
 
eca4fbf
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
Overview

SinBerto is a small language model trained on a small news corpus. SinBerto is trained on Sinhala Language which is a low resource language compared to other languages.

Model Specifications.

vocab_size=52_000,
max_position_embeddings=514,
num_attention_heads=12,
num_hidden_layers=6,
type_vocab_size=1,