snip-0.4m-base / tokenizer /tokenizer_config.json
ARotting's picture
Publish 397K parameter causal transformer pretrained from scratch
24ebd71 verified
Raw
History Blame Contribute Delete
204 Bytes
{
"backend": "tokenizers",
"bos_token": "<bos>",
"eos_token": "<eos>",
"model_max_length": 128,
"pad_token": "<pad>",
"tokenizer_class": "TokenizersBackend",
"unk_token": "<unk>"
}