{
  "architectures": [
    "GPTModel"
  ],
  "attention_probs_dropout_prob": 0.1,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "initializer_range": 0.02,
  "intermediate_size": 6400,
  "layer_norm_eps": 1e-12,
  "max_position_embeddings": 1024,
  "model_type": "gpt",
  "num_attention_heads": 16,
  "num_hidden_layers": 48,
  "hidden_size": 1600,
  "vocab_size": 50257,
  "summary_activation": "tanh",
  "summary_first_dropout": 0.1,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "bos_token_id": 50256,
  "eos_token_id": 50256,
  "pad_token_id": 0,
  "torch_dtype": "float32",
  "use_cache": true,
  "tokenizer_class": "GPT2Tokenizer"
}