Upload tokenizer
Browse files- added_tokens.json +14 -0
- special_tokens_map.json +20 -0
- spiece.model +3 -0
- tokenizer_config.json +11 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[book]": 11,
|
| 3 |
+
"[c4]": 12,
|
| 4 |
+
"[code]": 10,
|
| 5 |
+
"[convo]": 7,
|
| 6 |
+
"[eod]": 4,
|
| 7 |
+
"[eot]": 15,
|
| 8 |
+
"[fc]": 8,
|
| 9 |
+
"[ffc]": 9,
|
| 10 |
+
"[forum]": 14,
|
| 11 |
+
"[news]": 13,
|
| 12 |
+
"[translate]": 6,
|
| 13 |
+
"[wiki]": 5
|
| 14 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"[eod]",
|
| 4 |
+
"[wiki]",
|
| 5 |
+
"[translate]",
|
| 6 |
+
"[convo]",
|
| 7 |
+
"[fc]",
|
| 8 |
+
"[ffc]",
|
| 9 |
+
"[code]",
|
| 10 |
+
"[book]",
|
| 11 |
+
"[c4]",
|
| 12 |
+
"[news]",
|
| 13 |
+
"[forum]",
|
| 14 |
+
"[eot]"
|
| 15 |
+
],
|
| 16 |
+
"bos_token": "<s>",
|
| 17 |
+
"eos_token": "</s>",
|
| 18 |
+
"pad_token": "<pad>",
|
| 19 |
+
"unk_token": "<unk>"
|
| 20 |
+
}
|
spiece.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3909a67b780650b35cf529ac782ad2b6b26e6d1f849d3fbb6a872905f452458
|
| 3 |
+
size 4548313
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 6 |
+
"pad_token": "<pad>",
|
| 7 |
+
"sep_token": "<::::>",
|
| 8 |
+
"sp_model_kwargs": {},
|
| 9 |
+
"tokenizer_class": "BertGenerationTokenizer",
|
| 10 |
+
"unk_token": "<unk>"
|
| 11 |
+
}
|