hmm-tokenizer / tokenizer.json
fenguhao's picture
Upload tokenizer
08fe0fd verified
{
"version": "1.0",
"truncation": null,
"padding": null,
"added_tokens": [
{
"id": 0,
"content": "0 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 1,
"content": "1 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 2,
"content": "2 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 3,
"content": "3 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 4,
"content": "4 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 5,
"content": "5 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 6,
"content": "6 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 7,
"content": "7 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 8,
"content": "8 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 9,
"content": "9 ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 10,
"content": "a ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 11,
"content": "b ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 12,
"content": "c ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 13,
"content": "d ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 14,
"content": "e ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 15,
"content": "f ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 16,
"content": "g ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 17,
"content": "h ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 18,
"content": "i ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 19,
"content": "j ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 20,
"content": "k ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 21,
"content": "l ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 22,
"content": "m ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 23,
"content": "n ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 24,
"content": "o ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 25,
"content": "p ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 26,
"content": "q ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 27,
"content": "r ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 28,
"content": "s ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 29,
"content": "t ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 30,
"content": "u ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 31,
"content": "v ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 32,
"content": "w ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 33,
"content": "x ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 34,
"content": "y ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
},
{
"id": 35,
"content": "z ",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": true,
"special": false
}
],
"normalizer": null,
"pre_tokenizer": {
"type": "ByteLevel",
"add_prefix_space": false,
"trim_offsets": true,
"use_regex": true
},
"post_processor": {
"type": "ByteLevel",
"add_prefix_space": true,
"trim_offsets": false,
"use_regex": true
},
"decoder": {
"type": "ByteLevel",
"add_prefix_space": true,
"trim_offsets": true,
"use_regex": true
},
"model": {
"type": "BPE",
"dropout": null,
"unk_token": null,
"continuing_subword_prefix": "",
"end_of_word_suffix": "",
"fuse_unk": false,
"byte_fallback": false,
"ignore_merges": false,
"vocab": {
"0 ": 0
},
"merges": []
}
}