mirror of
https://github.com/zenlm/zen-family.git
synced 2026-11-03 14:26:45 +00:00
45 lines
998 B
JSON
45 lines
998 B
JSON
{
|
|
"model_max_length": 32768,
|
|
"tokenizer_class": "PreTrainedTokenizerFast",
|
|
"bos_token": "<s>",
|
|
"eos_token": "</s>",
|
|
"pad_token": "<pad>",
|
|
"unk_token": "<unk>",
|
|
"add_bos_token": true,
|
|
"add_eos_token": false,
|
|
"clean_up_tokenization_spaces": false,
|
|
"added_tokens_decoder": {
|
|
"50000": {
|
|
"content": "<|image|>",
|
|
"lstrip": false,
|
|
"normalized": false,
|
|
"rstrip": false,
|
|
"single_word": false,
|
|
"special": true
|
|
},
|
|
"50001": {
|
|
"content": "<|/image|>",
|
|
"lstrip": false,
|
|
"normalized": false,
|
|
"rstrip": false,
|
|
"single_word": false,
|
|
"special": true
|
|
},
|
|
"50002": {
|
|
"content": "<|audio|>",
|
|
"lstrip": false,
|
|
"normalized": false,
|
|
"rstrip": false,
|
|
"single_word": false,
|
|
"special": true
|
|
},
|
|
"50003": {
|
|
"content": "<|/audio|>",
|
|
"lstrip": false,
|
|
"normalized": false,
|
|
"rstrip": false,
|
|
"single_word": false,
|
|
"special": true
|
|
}
|
|
}
|
|
} |