{
  "attention_bias": false,
  "attention_dropout": 0.0,
  "attn_output_gate": true,
  "bos_token_id": 248044,
  "dtype": "bfloat16",
  "eos_token_id": 248046,
  "full_attention_interval": 4,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 64,
  "initializer_range": 0.02,
  "intermediate_size": 256,
  "layer_types": [
    "linear_attention",
    "linear_attention",
    "full_attention",
    "full_attention"
  ],
  "linear_conv_kernel_dim": 4,
  "linear_key_head_dim": 128,
  "linear_num_key_heads": 2,
  "linear_num_value_heads": 6,
  "linear_value_head_dim": 128,
  "mamba_ssm_dtype": "float32",
  "max_position_embeddings": 32768,
  "model_type": "volundr_text",
  "mtp_num_hidden_layers": 0,
  "mtp_use_dedicated_embeddings": false,
  "num_attention_heads": 2,
  "num_hidden_layers": 4,
  "num_key_value_heads": 1,
  "output_gate_type": "swish",
  "pad_token_id": 248044,
  "partial_rotary_factor": 0.25,
  "rms_norm_eps": 1e-06,
  "rope_parameters": {
    "mrope_interleaved": true,
    "mrope_section": [
      6,
      5,
      5
    ],
    "partial_rotary_factor": 0.25,
    "rope_theta": 10000000,
    "rope_type": "default"
  },
  "tie_word_embeddings": false,
  "use_cache": true,
  "vocab_size": 248320,
  "architectures": [
    "VolundrForCausalLM"
  ],
  "mixer_types": [
    "kda",
    "kda",
    "bcsa",
    "attn"
  ],
  "new_block_indices": [],
  "kda_gate_rank": 16,
  "bcsa_window": 64,
  "bcsa_compress": 4,
  "bcsa_topk": 8,
  "engram_layers": [
    1
  ],
  "mhc_streams": 4,
  "engram_orders": [
    2,
    3
  ],
  "engram_rows": 4099,
  "engram_dim": 32,
  "mhc_sinkhorn_iters": 10,
  "bcsa_impl": "ref"
}