{
  "processor_class": "Qwen2TokenizerFast",
  "tokenizer_class": "Qwen2Tokenizer",
  "do_normalize": false,
  "feature_extractor_type": null
}