gpt1_no-pretrain_xnli / tokenizer_config.json
tmnam20's picture
Training in progress, step 20000
d2a8393
raw
history blame
163 Bytes
{
"clean_up_tokenization_spaces": true,
"model_max_length": 512,
"padding_side": "left",
"tokenizer_class": "OpenAIGPTTokenizer",
"unk_token": "<unk>"
}