14 lines
236 B
JSON
14 lines
236 B
JSON
{
|
|
"vocab_size": 49152,
|
|
"d_model": 384,
|
|
"n_heads": 6,
|
|
"n_kv_heads": 2,
|
|
"n_layers": 24,
|
|
"d_ff": 1024,
|
|
"head_dim": 64,
|
|
"max_seq_len": 2048,
|
|
"rope_theta": 10000.0,
|
|
"rms_eps": 1e-05,
|
|
"qkv_bias": true,
|
|
"dropout": 0.0
|
|
} |