df-arc / tokenizer_config.json

Upload custom Unigram tokenizer (v1)

ea23987 verified 10 months ago

255 Bytes

{
  "tokenizer_class": "LlamaTokenizer",
  "model_max_length": 32768,
  "clean_up_tokenization_spaces": false,
  "unk_token": "[UNK]",
  "bos_token": "<s>",
  "eos_token": "</s>",
  "pad_token": "[PAD]",
  "add_bos_token": true,
  "add_eos_token": false
}

	{
	"tokenizer_class": "LlamaTokenizer",
	"model_max_length": 32768,
	"clean_up_tokenization_spaces": false,
	"unk_token": "[UNK]",
	"bos_token": "<s>",
	"eos_token": "</s>",
	"pad_token": "[PAD]",
	"add_bos_token": true,
	"add_eos_token": false
	}