tokenizer_config.json

{
  "bos_token": "<|startoftext|>",
  "eos_token": "<|endoftext|>",
  "model_max_length": 1000000000000000019884624838656,
  "pad_token": "<|pad|>",
  "special_tokens_map_file": "geclm/tokenizers/split_digits/all_data_concatenated/special_tokens_map.json",
  "tokenizer_class": "PreTrainedTokenizerFast",
  "unk_token": "<|unk|>"
}