{
  "model_type": "gpt2",
  "architectures": ["GPT2LMHeadModel"],
  "hidden_size": 768,
  "num_hidden_layers": 12,
  "num_attention_heads": 12,
  "vocab_size": 50257,
  "max_position_embeddings": 1024,
  "initializer_range": 0.02,
  "layer_norm_eps": 1e-5
}