config.json

{
  "architectures": [
    "HunYuanForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "attention_head_dim": 128,
  "auto_map": {
    "AutoConfig": "configuration_hunyuan.HunYuanConfig",
    "AutoModel": "modeling_hunyuan.HunyuanModel",
    "AutoModelForCausalLM": "modeling_hunyuan.HunYuanForCausalLM"
  },
  "bos_token_id": 1,
  "cla_share_factor": 2,
  "eos_token_id": 2,
  "group_limited_greedy": false,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "kv_lora_rank": null,
  "max_position_embeddings": 4096,
  "mlp_bias": false,
  "model_type": "hunyuan",
  "moe_drop_tokens": false,
  "moe_intermediate_size": [
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336,
    14336
  ],
  "moe_layer_num_skipped": 0,
  "moe_random_routing_dropped_token": false,
  "moe_topk": [
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1
  ],
  "n_group": false,
  "norm_topk_prob": false,
  "num_attention_heads": 32,
  "num_experts": 1,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "num_shared_expert": [
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1,
    1
  ],
  "pad_token_id": 0,
  "pretraining_tp": 1,
  "q_lora_rank": null,
  "qk_nope_head_dim": null,
  "qk_rope_head_dim": null,
  "rms_norm_eps": 1e-05,
  "rope_scaling": {
    "alpha": 1000.0,
    "beta_fast": 32,
    "beta_slow": 1,
    "factor": 1.0,
    "mscale": 1.0,
    "mscale_all_dim": 1.0,
    "type": "dynamic"
  },
  "rope_theta": 10000.0,
  "routed_scaling_factor": false,
  "tie_word_embeddings": true,
  "topk_group": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.41.2",
  "use_cache": true,
  "use_cla": false,
  "use_mixed_mlp_moe": false,
  "use_mla": false,
  "use_qk_norm": true,
  "v_head_dim": null,
  "vocab_size": 129024
}