tokenizer_config.json

{
  "vocab_size": 16000,
  "special_tokens": {
    "<|pad|>": 0,
    "<|eos|>": 1,
    "<|unk|>": 2,
    "<|system|>": 3,
    "<|user|>": 4,
    "<|assistant|>": 5,
    "<|code|>": 6,
    "<|comment|>": 7,
    "<|indent|>": 8,
    "<|newline|>": 9,
    "<|tab|>": 10,
    "<|space|>": 11
  }
}