Commit
•
88e3087
1
Parent(s):
dac3fb4
Improved tokenizer
Browse files- added_tokens.json +2 -1
- config.json +1 -1
- pytorch_model.bin +2 -2
- tokenizer.json +9 -0
- tokenizer_config.json +8 -0
- training_args.bin +1 -1
added_tokens.json
CHANGED
|
@@ -4,5 +4,6 @@
|
|
| 4 |
"[MASK]": 103,
|
| 5 |
"[PAD]": 0,
|
| 6 |
"[SEP]": 102,
|
| 7 |
-
"[UNK]": 100
|
|
|
|
| 8 |
}
|
|
|
|
| 4 |
"[MASK]": 103,
|
| 5 |
"[PAD]": 0,
|
| 6 |
"[SEP]": 102,
|
| 7 |
+
"[UNK]": 100,
|
| 8 |
+
"[URL]": 30523
|
| 9 |
}
|
config.json
CHANGED
|
@@ -32,5 +32,5 @@
|
|
| 32 |
"transformers_version": "4.34.0",
|
| 33 |
"type_vocab_size": 2,
|
| 34 |
"use_cache": true,
|
| 35 |
-
"vocab_size":
|
| 36 |
}
|
|
|
|
| 32 |
"transformers_version": "4.34.0",
|
| 33 |
"type_vocab_size": 2,
|
| 34 |
"use_cache": true,
|
| 35 |
+
"vocab_size": 30524
|
| 36 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81ad53d92c2e6e7974b64abf67be4dfcd32c80702ab9b205be5d544d8c139ccd
|
| 3 |
+
size 438132722
|
tokenizer.json
CHANGED
|
@@ -70,6 +70,15 @@
|
|
| 70 |
"rstrip": false,
|
| 71 |
"normalized": true,
|
| 72 |
"special": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 73 |
}
|
| 74 |
],
|
| 75 |
"normalizer": {
|
|
|
|
| 70 |
"rstrip": false,
|
| 71 |
"normalized": true,
|
| 72 |
"special": false
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"id": 30523,
|
| 76 |
+
"content": "[URL]",
|
| 77 |
+
"single_word": false,
|
| 78 |
+
"lstrip": false,
|
| 79 |
+
"rstrip": false,
|
| 80 |
+
"normalized": true,
|
| 81 |
+
"special": false
|
| 82 |
}
|
| 83 |
],
|
| 84 |
"normalizer": {
|
tokenizer_config.json
CHANGED
|
@@ -47,6 +47,14 @@
|
|
| 47 |
"rstrip": false,
|
| 48 |
"single_word": false,
|
| 49 |
"special": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
}
|
| 51 |
},
|
| 52 |
"additional_special_tokens": [],
|
|
|
|
| 47 |
"rstrip": false,
|
| 48 |
"single_word": false,
|
| 49 |
"special": false
|
| 50 |
+
},
|
| 51 |
+
"30523": {
|
| 52 |
+
"content": "[URL]",
|
| 53 |
+
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
+
"rstrip": false,
|
| 56 |
+
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
}
|
| 59 |
},
|
| 60 |
"additional_special_tokens": [],
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4472
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b9acb4e58e8ed6bf0bd605a87d385ad8d18dd39d69684d4df4033eb2f0e53238
|
| 3 |
size 4472
|