File size: 178 Bytes
d79b2b9
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
{
    n_layers = 6,
    d_model = 128,
    d_head = 64,
    n_heads = 8,
    d_mlp = 512,
    d_vocab = 61,
    n_ctx = 59,
    act_fn="gelu",
    normalization_type="LNPre",
  }