{
  "architectures": [
    "ZayaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 2,
  "cca_time0": 2,
  "cca_time1": 2,
  "dtype": "bfloat16",
  "eos_token_id": 106,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "layer_types": [
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid",
    "hybrid"
  ],
  "lm_head_bias": false,
  "max_position_embeddings": 131072,
  "model_type": "zaya",
  "moe_intermediate_size": 2048,
  "num_attention_heads": 8,
  "num_experts": 16,
  "num_experts_per_tok": 1,
  "num_hidden_layers": 40,
  "num_key_value_heads": 2,
  "output_router_logits": false,
  "pad_token_id": 0,
  "partial_rotary_factor": 0.5,
  "rms_norm_eps": 1e-05,
  "rope_parameters": {
    "hybrid": {
      "partial_rotary_factor": 0.5,
      "rope_theta": 5000000,
      "rope_type": "default"
    },
    "hybrid_sliding": {
      "partial_rotary_factor": 0.5,
      "rope_theta": 10000.0,
      "rope_type": "default"
    },
    "rope_type": "default"
  },
  "router_hidden_size": 256,
  "sliding_window": null,
  "tie_word_embeddings": true,
  "transformers_version": "5.10.0.dev0",
  "use_cache": true,
  "vocab_size": 262272
}