config.json

{
  "architectures": [
    "LingoWhaleForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "configuration_lingowhale.LingoWhaleConfig",
    "AutoModelForCausalLM": "modeling_lingowhale.LingoWhaleForCausalLM"
  },
  "tokenizer_class": "LingoWhaleTokenizer",
  "bos_token_id": 1,
  "eos_token_id": 2,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 11008,
  "max_position_embeddings": 8192,
  "model_max_length": 8192,
  "model_type": "lingowhale",
  "num_attention_heads": 32,
  "num_hidden_layers": 36,
  "pad_token_id": 0,
  "rms_norm_eps": 1e-06,
  "_from_model_config": true,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.31.0",
  "use_cache": true,
  "emb_dropout_prob": 0.0,
  "attn_dropout_prob": 0.0,
  "vocab_size": 96000,
  "rope_theta": 10000.0,
  "use_flash_attention": true
}