Upload metre-char-tokenizer.json with huggingface_hub
Browse files- metre-char-tokenizer.json +62 -0
metre-char-tokenizer.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": {
|
| 5 |
+
"strategy": "BatchLongest",
|
| 6 |
+
"direction": "Right",
|
| 7 |
+
"pad_to_multiple_of": null,
|
| 8 |
+
"pad_id": 0,
|
| 9 |
+
"pad_type_id": 0,
|
| 10 |
+
"pad_token": "[PAD]"
|
| 11 |
+
},
|
| 12 |
+
"added_tokens": [
|
| 13 |
+
{
|
| 14 |
+
"id": 0,
|
| 15 |
+
"content": "[PAD]",
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"lstrip": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"id": 1,
|
| 24 |
+
"content": "[SOS]",
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"lstrip": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"normalized": false,
|
| 29 |
+
"special": true
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"id": 2,
|
| 33 |
+
"content": "[EOS]",
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"lstrip": false,
|
| 36 |
+
"rstrip": false,
|
| 37 |
+
"normalized": false,
|
| 38 |
+
"special": true
|
| 39 |
+
}
|
| 40 |
+
],
|
| 41 |
+
"normalizer": null,
|
| 42 |
+
"pre_tokenizer": null,
|
| 43 |
+
"post_processor": null,
|
| 44 |
+
"decoder": null,
|
| 45 |
+
"model": {
|
| 46 |
+
"type": "BPE",
|
| 47 |
+
"dropout": null,
|
| 48 |
+
"unk_token": null,
|
| 49 |
+
"continuing_subword_prefix": null,
|
| 50 |
+
"end_of_word_suffix": null,
|
| 51 |
+
"fuse_unk": false,
|
| 52 |
+
"byte_fallback": false,
|
| 53 |
+
"vocab": {
|
| 54 |
+
"[PAD]": 0,
|
| 55 |
+
"[SOS]": 1,
|
| 56 |
+
"[EOS]": 2,
|
| 57 |
+
"-": 3,
|
| 58 |
+
"U": 4
|
| 59 |
+
},
|
| 60 |
+
"merges": []
|
| 61 |
+
}
|
| 62 |
+
}
|