tokenizer_config.json

{
  "added_tokens_decoder": {
    "32759": {
      "content": "<|bos|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32760": {
      "content": "<|user_start|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32761": {
      "content": "<|user_end|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32762": {
      "content": "<|assistant_start|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32763": {
      "content": "<|assistant_end|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32764": {
      "content": "<|python_start|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32765": {
      "content": "<|python_end|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32766": {
      "content": "<|output_start|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "32767": {
      "content": "<|output_end|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    }
  },
  "additional_special_tokens": [
    "<|user_start|>",
    "<|user_end|>",
    "<|assistant_start|>",
    "<|assistant_end|>",
    "<|python_start|>",
    "<|python_end|>",
    "<|output_start|>",
    "<|output_end|>"
  ],
  "bos_token": "<|bos|>",
  "clean_up_tokenization_spaces": false,
  "eos_token": "<|bos|>",
  "extra_special_tokens": {},
  "model_input_names": [
    "input_ids",
    "attention_mask"
  ],
  "model_max_length": 2048,
  "pad_token": "<|bos|>",
  "tokenizer_class": "PreTrainedTokenizerFast",
  "unk_token": null
}