v76-34kdataset-512k-bs343-2te-69k-bs10-170k-768-cd0.02-72k-rebal2-w-te-cd0.001-37k-halfn
/
tokenizer
/tokenizer_config.json
{ | |
"add_prefix_space": false, | |
"added_tokens_decoder": { | |
"0": { | |
"content": "!", | |
"lstrip": false, | |
"normalized": false, | |
"rstrip": false, | |
"single_word": false, | |
"special": true | |
}, | |
"49406": { | |
"content": "<|startoftext|>", | |
"lstrip": false, | |
"normalized": true, | |
"rstrip": false, | |
"single_word": false, | |
"special": true | |
}, | |
"49407": { | |
"content": "<|endoftext|>", | |
"lstrip": false, | |
"normalized": true, | |
"rstrip": false, | |
"single_word": false, | |
"special": true | |
} | |
}, | |
"bos_token": "<|startoftext|>", | |
"clean_up_tokenization_spaces": true, | |
"do_lower_case": true, | |
"eos_token": "<|endoftext|>", | |
"errors": "replace", | |
"model_max_length": 77, | |
"pad_token": "!", | |
"tokenizer_class": "CLIPTokenizer", | |
"unk_token": "<|endoftext|>", | |
"use_fast": false | |
} | |