Upload tokenizer
Browse files- added_tokens.json +4 -0
- special_tokens_map.json +5 -3
- tokenizer_config.json +1 -7
added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[DEC]": 30522,
|
| 3 |
+
"[ENC]": 30523
|
| 4 |
+
}
|
special_tokens_map.json
CHANGED
|
@@ -1,9 +1,11 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
"cls_token": "[CLS]",
|
| 3 |
"mask_token": "[MASK]",
|
| 4 |
"pad_token": "[PAD]",
|
| 5 |
"sep_token": "[SEP]",
|
| 6 |
-
"unk_token": "[UNK]"
|
| 7 |
-
"bos_token": "[DEC]",
|
| 8 |
-
"additional_special_tokens": "[ENC]"
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"[ENC]"
|
| 4 |
+
],
|
| 5 |
+
"bos_token": "[DEC]",
|
| 6 |
"cls_token": "[CLS]",
|
| 7 |
"mask_token": "[MASK]",
|
| 8 |
"pad_token": "[PAD]",
|
| 9 |
"sep_token": "[SEP]",
|
| 10 |
+
"unk_token": "[UNK]"
|
|
|
|
|
|
|
| 11 |
}
|
tokenizer_config.json
CHANGED
|
@@ -3,16 +3,10 @@
|
|
| 3 |
"do_basic_tokenize": true,
|
| 4 |
"do_lower_case": true,
|
| 5 |
"mask_token": "[MASK]",
|
| 6 |
-
"model_input_names": [
|
| 7 |
-
"input_ids",
|
| 8 |
-
"attention_mask"
|
| 9 |
-
],
|
| 10 |
"model_max_length": 512,
|
| 11 |
-
"name_or_path": "
|
| 12 |
"never_split": null,
|
| 13 |
"pad_token": "[PAD]",
|
| 14 |
-
"bos_token": "[DEC]",
|
| 15 |
-
"processor_class": "BlipProcessor",
|
| 16 |
"sep_token": "[SEP]",
|
| 17 |
"special_tokens_map_file": null,
|
| 18 |
"strip_accents": null,
|
|
|
|
| 3 |
"do_basic_tokenize": true,
|
| 4 |
"do_lower_case": true,
|
| 5 |
"mask_token": "[MASK]",
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
"model_max_length": 512,
|
| 7 |
+
"name_or_path": "bert-base-uncased",
|
| 8 |
"never_split": null,
|
| 9 |
"pad_token": "[PAD]",
|
|
|
|
|
|
|
| 10 |
"sep_token": "[SEP]",
|
| 11 |
"special_tokens_map_file": null,
|
| 12 |
"strip_accents": null,
|