NousResearch

/

Hermes-2-Theta-Llama-3-8B

like 167

Link: https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B/commit/885173e97ab8572b444f7db1290d5d0386e26816

  },
    {
      "id": 128005,
      "content": "<|reserved_special_token_3|>",
      "single_word": false,
      "lstrip": false,
      "rstrip": false,

    "end_of_word_suffix": null,
    "fuse_unk": false,
    "byte_fallback": false,
    "ignore_merges": true,
    "vocab": {
      "!": 0,
      "\"": 1,

Can you tell me why <tool_response> is changed to <|reserved_special_token_3|>?
I am just practicing the model merging.
After the tokenizer.json file is changed to the current version, it always meets the tokenization error.