{
  "activation_type": "silu",
  "architectures": [
    "AliceAIT5MoEForConditionalGeneration"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_aliceai_t5_moe.AliceAIT5MoEConfig",
    "AutoModel": "modeling_aliceai_t5_moe.AliceAIT5MoEModel",
    "AutoModelForSeq2SeqLM": "modeling_aliceai_t5_moe.AliceAIT5MoEForConditionalGeneration"
  },
  "bos_token_id": 1,
  "decoder": {
    "attention_bias": false,
    "bos_token_id": 1,
    "fp32_residual": true,
    "head_dim": 128,
    "hidden_size": 1536,
    "initializer_range": 0.02,
    "max_position_embeddings": 131073,
    "model_type": "aliceai_t5_module",
    "norm_eps": 1e-06,
    "num_attention_heads": 12,
    "num_hidden_layers": 12,
    "num_key_value_heads": 4,
    "pad_token_id": 2,
    "query_pre_attn_scalar": 128,
    "rope_parameters": {
      "attention_factor": 1.299573227355399,
      "factor": 20.0,
      "original_max_position_embeddings": 9984,
      "rope_type": "yarn"
    },
    "rope_theta": 10000.0,
    "sliding_window": null,
    "use_cache": true,
    "vocab_size": 135040
  },
  "dropout_rate": 0.0,
  "dtype": "bfloat16",
  "encoder": {
    "attention_bias": false,
    "bos_token_id": 1,
    "fp32_residual": true,
    "head_dim": 128,
    "hidden_size": 1536,
    "initializer_range": 0.02,
    "max_position_embeddings": 131073,
    "model_type": "aliceai_t5_module",
    "norm_eps": 1e-06,
    "num_attention_heads": 12,
    "num_hidden_layers": 16,
    "num_key_value_heads": 12,
    "pad_token_id": 2,
    "query_pre_attn_scalar": 128,
    "rope_parameters": {
      "attention_factor": 1.299573227355399,
      "factor": 20.0,
      "original_max_position_embeddings": 9984,
      "rope_type": "yarn"
    },
    "rope_theta": 10000.0,
    "sliding_window": null,
    "use_cache": true,
    "vocab_size": 135040
  },
  "eos_token_id": 2,
  "expert_intermediate_size": 512,
  "experts_gating_top": 8,
  "experts_n_group": 1,
  "experts_top_k_group": 1,
  "group_routing": false,
  "initializer_range": 0.02,
  "is_encoder_decoder": true,
  "model_type": "aliceai_t5_moe",
  "moe_normalize_expert_weights": true,
  "n_experts": 512,
  "pad_token_id": 2,
  "routed_scaling_factor": 1.0,
  "shared_embeddings": true,
  "tie_word_embeddings": true,
  "transformers_version": "5.14.1",
  "use_cache": true
}