cogent-csp-1m / tokenizer_config.json
pure-team's picture
Upload 5 files
a5617bb verified
raw
history blame
227 Bytes
{
"model_max_length": 128,
"tokenizer_class": "PreTrainedTokenizerFast",
"padding_side": "left",
"truncation_side": "right",
"bos_token": "<s>",
"eos_token": "</s>",
"unk_token": "<unk>",
"pad_token": "<pad>"
}