{
  "_name_or_path": "./results/Doge-20M-Instruct-DPO",
  "architectures": [
    "DogeForCausalLM"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_doge.DogeConfig",
    "AutoModelForCausalLM": "modeling_doge.DogeForCausalLM"
  },
  "bos_token_id": 0,
  "dynamic_mask_ratio": 0.0,
  "eos_token_id": 1,
  "expert_retrieval_size": 256,
  "hidden_act": "silu",
  "hidden_bias": false,
  "hidden_dropout": 0.0,
  "hidden_size": 256,
  "initializer_range": 0.02,
  "intermediate_size": 512,
  "is_moe": false,
  "max_position_embeddings": 2048,
  "model_type": "doge",
  "num_attention_heads": 2,
  "num_cdmmoe_experts": 2048,
  "num_cdmmoe_experts_per_head": 8,
  "num_cdmmoe_heads": 4,
  "num_channels": 3,
  "num_hidden_layers": 8,
  "num_key_value_heads": 1,
  "pad_token_id": 2,
  "patch_size": 16,
  "rms_norm_eps": 1e-06,
  "rope_scaling": {
    "factor": 4.0,
    "original_max_position_embeddings": 2048,
    "rope_type": "dynamic"
  },
  "rope_theta": 10000.0,
  "torch_dtype": "float32",
  "transformers_version": "4.49.0.dev0",
  "use_cache": true,
  "vocab_size": 32768
}