Upload tokenizer
Browse files- tokenizer_config.json +5 -0
tokenizer_config.json
CHANGED
|
@@ -56,12 +56,17 @@
|
|
| 56 |
"eos_token": "<|endoftext|>",
|
| 57 |
"errors": "replace",
|
| 58 |
"mask_token": "<mask>",
|
|
|
|
| 59 |
"model_max_length": 2048,
|
|
|
|
| 60 |
"pad_token": "<pad>",
|
|
|
|
| 61 |
"padding_side": "right",
|
| 62 |
"sep_token": "</s>",
|
|
|
|
| 63 |
"tokenizer_class": "GPT2Tokenizer",
|
| 64 |
"truncation_side": "left",
|
|
|
|
| 65 |
"trust_remote_code": false,
|
| 66 |
"unk_token": "<unk>"
|
| 67 |
}
|
|
|
|
| 56 |
"eos_token": "<|endoftext|>",
|
| 57 |
"errors": "replace",
|
| 58 |
"mask_token": "<mask>",
|
| 59 |
+
"max_length": 200,
|
| 60 |
"model_max_length": 2048,
|
| 61 |
+
"pad_to_multiple_of": null,
|
| 62 |
"pad_token": "<pad>",
|
| 63 |
+
"pad_token_type_id": 0,
|
| 64 |
"padding_side": "right",
|
| 65 |
"sep_token": "</s>",
|
| 66 |
+
"stride": 0,
|
| 67 |
"tokenizer_class": "GPT2Tokenizer",
|
| 68 |
"truncation_side": "left",
|
| 69 |
+
"truncation_strategy": "longest_first",
|
| 70 |
"trust_remote_code": false,
|
| 71 |
"unk_token": "<unk>"
|
| 72 |
}
|