Commit
·
2ccea53
1
Parent(s):
dd42733
Upload tokenizer
Browse files- tokenizer.json +2 -16
- tokenizer_config.json +7 -0
tokenizer.json
CHANGED
|
@@ -1,21 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
|
| 5 |
-
"max_length": 48,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
-
"padding": {
|
| 10 |
-
"strategy": {
|
| 11 |
-
"Fixed": 48
|
| 12 |
-
},
|
| 13 |
-
"direction": "Right",
|
| 14 |
-
"pad_to_multiple_of": null,
|
| 15 |
-
"pad_id": 50256,
|
| 16 |
-
"pad_type_id": 0,
|
| 17 |
-
"pad_token": "<|endoftext|>"
|
| 18 |
-
},
|
| 19 |
"added_tokens": [
|
| 20 |
{
|
| 21 |
"id": 0,
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
| 7 |
"id": 0,
|
tokenizer_config.json
CHANGED
|
@@ -108,8 +108,15 @@
|
|
| 108 |
"eos_token": "<|endoftext|>",
|
| 109 |
"errors": "replace",
|
| 110 |
"full_tokenizer_file": null,
|
|
|
|
| 111 |
"model_max_length": 1000000000000000019884624838656,
|
|
|
|
| 112 |
"pad_token": "<|endoftext|>",
|
|
|
|
|
|
|
|
|
|
| 113 |
"tokenizer_class": "GPT2Tokenizer",
|
|
|
|
|
|
|
| 114 |
"unk_token": "<|endoftext|>"
|
| 115 |
}
|
|
|
|
| 108 |
"eos_token": "<|endoftext|>",
|
| 109 |
"errors": "replace",
|
| 110 |
"full_tokenizer_file": null,
|
| 111 |
+
"max_length": 48,
|
| 112 |
"model_max_length": 1000000000000000019884624838656,
|
| 113 |
+
"pad_to_multiple_of": null,
|
| 114 |
"pad_token": "<|endoftext|>",
|
| 115 |
+
"pad_token_type_id": 0,
|
| 116 |
+
"padding_side": "right",
|
| 117 |
+
"stride": 0,
|
| 118 |
"tokenizer_class": "GPT2Tokenizer",
|
| 119 |
+
"truncation_side": "right",
|
| 120 |
+
"truncation_strategy": "longest_first",
|
| 121 |
"unk_token": "<|endoftext|>"
|
| 122 |
}
|