tokenizer_config.json

{
  "add_prefix_space": false,
  "added_tokens_decoder": {
    "0": {
      "content": "<|endofmolecule|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "4600": {
      "content": "<|startofmolecule|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    },
    "4601": {
      "content": "<|UNK|>",
      "lstrip": false,
      "normalized": false,
      "rstrip": false,
      "single_word": false,
      "special": true
    }
  },
  "additional_special_tokens": [
    "<|endofmolecule|>"
  ],
  "bos_token": "<|startofmolecule|>",
  "clean_up_tokenization_spaces": true,
  "eos_token": "<|endofmolecule|>",
  "model_max_length": 1024,
  "tokenizer_class": "GPT2Tokenizer",
  "unk_token": "<|UNK|>"
}