{
  "add_attn_pool_reduce_stream": false,
  "add_stream_embed": true,
  "architectures": [
    "Qwen3MHCForCausalLMV2"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "dtype": "bfloat16",
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 1024,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "layer_types": [
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention",
    "full_attention"
  ],
  "log_domain_sinkhorn": false,
  "max_position_embeddings": 40960,
  "max_window_layers": 28,
  "mhc_dropout": 0.0,
  "mlp_bias": false,
  "model_type": "qwen3_mhc_v2",
  "n_streams": 4,
  "num_attention_heads": 16,
  "num_dynamic_alpha_proposals": 1,
  "num_fracs": 1,
  "num_hidden_layers": 28,
  "num_input_views": 1,
  "num_key_value_heads": 8,
  "residual_mix_temperature": 1.0,
  "rms_norm_eps": 1e-06,
  "rope_parameters": {
    "rope_theta": 1000000,
    "rope_type": "default"
  },
  "rope_theta": 1000000.0,
  "sinkhorn_iters": 20,
  "sliding_window": 4096,
  "tie_word_embeddings": true,
  "transformers_version": "5.0.0",
  "use_cache": false,
  "use_sliding_window": false,
  "use_triton_sinkhorn": false,
  "vocab_size": 151936
}