tokenizer_config.json

{
  "model_type": "gpt2",
  "vocab_size": 50257,
  "padding_side": "right",
  "truncation_side": "right",
  "bos_token_id": 50256,
  "eos_token_id": 50256
}