config.json

{
  "architectures": [
    "Qwen3DSparkModel"
  ],
  "model_type": "qwen3",
  "block_size": 8,
  "dflash_config": {
    "mask_token_id": 200063,
    "target_layer_ids": [
      1,
      12,
      23,
      35,
      46,
      57
    ],
    "use_swa": true,
    "swa_window_size": 1024,
    "causal": false,
    "projector_type": "dspark",
    "shift_label": true,
    "markov_rank": 256,
    "markov_head_type": "vanilla",
    "use_confidence_head": true
  },
  "hidden_size": 6144,
  "num_hidden_layers": 6,
  "num_attention_heads": 32,
  "num_key_value_heads": 8,
  "head_dim": 128,
  "intermediate_size": 12288,
  "hidden_act": "silu",
  "rms_norm_eps": 1e-06,
  "vocab_size": 200064,
  "max_position_embeddings": 1048576,
  "initializer_range": 0.02,
  "attention_bias": false,
  "attention_dropout": 0.0,
  "rope_theta": 5000000,
  "rope_scaling": null,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "num_target_layers": 60,
  "layer_types": [
    "sliding_attention",
    "sliding_attention",
    "sliding_attention",
    "sliding_attention",
    "sliding_attention",
    "sliding_attention"
  ],
  "sliding_window": 1024,
  "markov_rank": 256,
  "markov_head_type": "vanilla"
}