tied-per-target-layerwise-token-topk/cfg.json

{
    "num_features": 32768,
    "num_layers": 12,
    "d_model": 768,
    "activation_fn": "topk",
    "topk_k": 16,
    "topk_straight_through": true,
    "topk_mode": "per_layer",
    "decoder_tying": "per_target",
    "enable_feature_offset": false,
    "enable_feature_scale": false,
    "skip_connection": false,
    "normalization_method": "mean_std"
  }