anlp-a2-tokenizers / tokenizer_stats.json
usrnotfound101's picture
tokenizers
fc83f97 verified
Raw History Blame Contribute Delete
828 Bytes
{
"mt_en": {
"mean_tokens": 18.8248,
"p95_tokens": 50.0,
"max_tokens": 112,
"tokens_per_char": 0.2717304496837542,
"tokens_per_whitespace_word": 1.5909059147501896,
"unk_tokens": 0,
"byte_fallback_tokens": 71
},
"mt_vi": {
"mean_tokens": 22.5454,
"p95_tokens": 60.0,
"max_tokens": 120,
"tokens_per_char": 0.30452669132286336,
"tokens_per_whitespace_word": 1.3601911156958297,
"unk_tokens": 0,
"byte_fallback_tokens": 105
},
"mt_ja": {
"mean_tokens": 17.9854,
"p95_tokens": 47.0,
"max_tokens": 110,
"tokens_per_char": 0.48894536698039154,
"tokens_per_whitespace_word": 10.451642071573078,
"unk_tokens": 0,
"byte_fallback_tokens": 1100
},
"lm_tokens_per_split": {
"train": 40141738,
"val": 2217353,
"test": 2248878
}
}