barun_config.json

{
  "activation_clip": 10.0,
  "attention_gate": true,
  "dim": 448,
  "dropout": 0.0,
  "ffn_dim": 1228,
  "full_attention_every": 4,
  "local_window": 256,
  "max_seq_len": 2048,
  "mtp_loss_weight": 0.0,
  "mtp_offset": 2,
  "n_heads": 7,
  "n_kv_heads": 1,
  "n_layers": 12,
  "norm_eps": 1e-06,
  "qk_norm": true,
  "residual_select_every": 4,
  "rope_fraction": 0.5,
  "rope_theta": 10000.0,
  "tie_embeddings": true,
  "vocab_size": 16384
}