config.json

{
  "architectures": [
    "HybridQwen3ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "dtype": "bfloat16",
  "eos_token_id": 151645,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "hybrid_override_pattern": "*-M2-M2-M2-M2-M2-M2-*-*-M2-M2-M2-*-*-*-*-*-*-*-*-*-*-*-*-*-*-M2-M2-M2-*-M2-M2-M2-M2-M2-M2",
  "initializer_range": 0.02,
  "intermediate_size": 12288,
  "layer_types": [
    "*",
    "M2",
    "M2",
    "M2",
    "M2",
    "M2",
    "M2",
    "*",
    "*",
    "M2",
    "M2",
    "M2",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "*",
    "M2",
    "M2",
    "M2",
    "*",
    "M2",
    "M2",
    "M2",
    "M2",
    "M2",
    "M2"
  ],
  "mamba2_config": {
    "d_inner": 4096,
    "d_model": 4096,
    "d_xb": 1024,
    "hidden_act": "silu",
    "intermediate_size": 12288,
    "n_layer": 36,
    "rms_norm_eps": 1e-06,
    "ssm_cfg": {
      "d_state": 128,
      "expand": 1,
      "ngroups": 32
    },
    "use_pos_emb": false,
    "use_qk_norm": true
  },
  "max_position_embeddings": 131072,
  "max_window_layers": 36,
  "model_type": "hybrid_qwen3",
  "num_attention_heads": 32,
  "num_hidden_layers": 36,
  "num_key_value_heads": 8,
  "pad_token_id": null,
  "rms_norm_eps": 1e-06,
  "rope_parameters": {
    "rope_theta": 5000000,
    "rope_type": "default"
  },
  "sliding_window": null,
  "tie_word_embeddings": false,
  "transformers_version": "5.3.0",
  "use_cache": false,
  "use_sliding_window": false,
  "vocab_size": 151936
}