Upload processor
Browse files- preprocessor_config.json +4 -4
- tokenizer.json +0 -0
- tokenizer_config.json +7 -0
preprocessor_config.json
CHANGED
|
@@ -37,8 +37,8 @@
|
|
| 37 |
"processor_class": "DonutProcessor",
|
| 38 |
"resample": 2,
|
| 39 |
"rescale_factor": 0.00392156862745098,
|
| 40 |
-
"size":
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
}
|
|
|
|
| 37 |
"processor_class": "DonutProcessor",
|
| 38 |
"resample": 2,
|
| 39 |
"rescale_factor": 0.00392156862745098,
|
| 40 |
+
"size": {
|
| 41 |
+
"height": 1280,
|
| 42 |
+
"width": 960
|
| 43 |
+
}
|
| 44 |
}
|
tokenizer.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
CHANGED
|
@@ -18834,11 +18834,18 @@
|
|
| 18834 |
"cls_token": "<s>",
|
| 18835 |
"eos_token": "</s>",
|
| 18836 |
"mask_token": "<mask>",
|
|
|
|
| 18837 |
"model_max_length": 1000000000000000019884624838656,
|
|
|
|
| 18838 |
"pad_token": "<pad>",
|
|
|
|
|
|
|
| 18839 |
"processor_class": "DonutProcessor",
|
| 18840 |
"sep_token": "</s>",
|
| 18841 |
"sp_model_kwargs": {},
|
|
|
|
| 18842 |
"tokenizer_class": "XLMRobertaTokenizer",
|
|
|
|
|
|
|
| 18843 |
"unk_token": "<unk>"
|
| 18844 |
}
|
|
|
|
| 18834 |
"cls_token": "<s>",
|
| 18835 |
"eos_token": "</s>",
|
| 18836 |
"mask_token": "<mask>",
|
| 18837 |
+
"max_length": 768,
|
| 18838 |
"model_max_length": 1000000000000000019884624838656,
|
| 18839 |
+
"pad_to_multiple_of": null,
|
| 18840 |
"pad_token": "<pad>",
|
| 18841 |
+
"pad_token_type_id": 0,
|
| 18842 |
+
"padding_side": "right",
|
| 18843 |
"processor_class": "DonutProcessor",
|
| 18844 |
"sep_token": "</s>",
|
| 18845 |
"sp_model_kwargs": {},
|
| 18846 |
+
"stride": 0,
|
| 18847 |
"tokenizer_class": "XLMRobertaTokenizer",
|
| 18848 |
+
"truncation_side": "right",
|
| 18849 |
+
"truncation_strategy": "longest_first",
|
| 18850 |
"unk_token": "<unk>"
|
| 18851 |
}
|