G0-nano-instruct / tokenizer_config.json
AZERDSQ's picture
Renomme les identifiants nanogpt -> g0nano (config, model, tokenizer)
8eb0d18 verified
Raw
History Blame Contribute Delete
389 Bytes
{
"tokenizer_class": "G0NanoTokenizer",
"auto_map": {
"AutoTokenizer": [
"tokenization_g0nano.G0NanoTokenizer",
null
]
},
"bos_token": "<s>",
"eos_token": "</s>",
"unk_token": "<unk>",
"pad_token": "<pad>",
"additional_special_tokens": [
"<|user|>",
"<|assistant|>",
"<|end|>",
"<|system|>"
],
"clean_up_tokenization_spaces": false
}