aptagpt_internal / tokenizer_config.json
sizhennnn's picture
Training in progress, step 838
6b10979 verified
{
"added_tokens_decoder": {
"1024": {
"content": "BOS",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1025": {
"content": "EOS",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1026": {
"content": "UNK",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1027": {
"content": "PAD",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
}
},
"bos_token": "BOS",
"clean_up_tokenization_spaces": true,
"eos_token": "EOS",
"model_max_length": 15,
"pad_token": "PAD",
"padding_side": "right",
"tokenizer_class": "PreTrainedTokenizerFast",
"truncation_side": "right",
"unk_token": "UNK"
}