Upload processor
Browse files- added_tokens.json +10 -9
- preprocessor_config.json +20 -2
- special_tokens_map.json +8 -1
- tokenizer.json +0 -0
- tokenizer_config.json +30 -15
added_tokens.json
CHANGED
|
@@ -1,11 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"</
|
| 3 |
-
"
|
| 4 |
-
"<
|
| 5 |
-
"<
|
| 6 |
-
"<
|
| 7 |
-
"<s_iitcdip>":
|
| 8 |
-
"<
|
| 9 |
-
"<s_synthdog>":
|
| 10 |
-
"<sep/>": 57522
|
|
|
|
| 11 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s_answer>": 57523,
|
| 3 |
+
"</s_question>": 57524,
|
| 4 |
+
"<no/>": 57525,
|
| 5 |
+
"<s_answer>": 57526,
|
| 6 |
+
"<s_docvqa>": 57527,
|
| 7 |
+
"<s_iitcdip>": 57528,
|
| 8 |
+
"<s_question>": 57529,
|
| 9 |
+
"<s_synthdog>": 57530,
|
| 10 |
+
"<sep/>": 57522,
|
| 11 |
+
"<yes/>": 57531
|
| 12 |
}
|
preprocessor_config.json
CHANGED
|
@@ -1,4 +1,22 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
"do_align_long_axis": false,
|
| 3 |
"do_normalize": true,
|
| 4 |
"do_pad": true,
|
|
@@ -20,7 +38,7 @@
|
|
| 20 |
"resample": 2,
|
| 21 |
"rescale_factor": 0.00392156862745098,
|
| 22 |
"size": [
|
| 23 |
-
|
| 24 |
-
|
| 25 |
]
|
| 26 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"_valid_processor_keys": [
|
| 3 |
+
"images",
|
| 4 |
+
"do_resize",
|
| 5 |
+
"size",
|
| 6 |
+
"resample",
|
| 7 |
+
"do_thumbnail",
|
| 8 |
+
"do_align_long_axis",
|
| 9 |
+
"do_pad",
|
| 10 |
+
"random_padding",
|
| 11 |
+
"do_rescale",
|
| 12 |
+
"rescale_factor",
|
| 13 |
+
"do_normalize",
|
| 14 |
+
"image_mean",
|
| 15 |
+
"image_std",
|
| 16 |
+
"return_tensors",
|
| 17 |
+
"data_format",
|
| 18 |
+
"input_data_format"
|
| 19 |
+
],
|
| 20 |
"do_align_long_axis": false,
|
| 21 |
"do_normalize": true,
|
| 22 |
"do_pad": true,
|
|
|
|
| 38 |
"resample": 2,
|
| 39 |
"rescale_factor": 0.00392156862745098,
|
| 40 |
"size": [
|
| 41 |
+
1920,
|
| 42 |
+
2560
|
| 43 |
]
|
| 44 |
}
|
special_tokens_map.json
CHANGED
|
@@ -1,7 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
"<s_iitcdip>",
|
| 4 |
-
"<
|
|
|
|
|
|
|
| 5 |
],
|
| 6 |
"bos_token": {
|
| 7 |
"content": "<s>",
|
|
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
| 3 |
+
"</s_answer>",
|
| 4 |
+
"</s_question>",
|
| 5 |
+
"<no/>",
|
| 6 |
+
"<s_answer>",
|
| 7 |
+
"<s_docvqa>",
|
| 8 |
"<s_iitcdip>",
|
| 9 |
+
"<s_question>",
|
| 10 |
+
"<s_synthdog>",
|
| 11 |
+
"<yes/>"
|
| 12 |
],
|
| 13 |
"bos_token": {
|
| 14 |
"content": "<s>",
|
tokenizer.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
CHANGED
|
@@ -49,23 +49,23 @@
|
|
| 49 |
"special": false
|
| 50 |
},
|
| 51 |
"57523": {
|
| 52 |
-
"content": "
|
| 53 |
"lstrip": false,
|
| 54 |
-
"normalized":
|
| 55 |
"rstrip": false,
|
| 56 |
"single_word": false,
|
| 57 |
-
"special":
|
| 58 |
},
|
| 59 |
"57524": {
|
| 60 |
-
"content": "
|
| 61 |
"lstrip": false,
|
| 62 |
-
"normalized":
|
| 63 |
"rstrip": false,
|
| 64 |
"single_word": false,
|
| 65 |
-
"special":
|
| 66 |
},
|
| 67 |
"57525": {
|
| 68 |
-
"content": "<
|
| 69 |
"lstrip": false,
|
| 70 |
"normalized": true,
|
| 71 |
"rstrip": false,
|
|
@@ -73,7 +73,7 @@
|
|
| 73 |
"special": false
|
| 74 |
},
|
| 75 |
"57526": {
|
| 76 |
-
"content": "<
|
| 77 |
"lstrip": false,
|
| 78 |
"normalized": true,
|
| 79 |
"rstrip": false,
|
|
@@ -81,7 +81,7 @@
|
|
| 81 |
"special": false
|
| 82 |
},
|
| 83 |
"57527": {
|
| 84 |
-
"content": "<
|
| 85 |
"lstrip": false,
|
| 86 |
"normalized": true,
|
| 87 |
"rstrip": false,
|
|
@@ -89,15 +89,15 @@
|
|
| 89 |
"special": false
|
| 90 |
},
|
| 91 |
"57528": {
|
| 92 |
-
"content": "<
|
| 93 |
"lstrip": false,
|
| 94 |
-
"normalized":
|
| 95 |
"rstrip": false,
|
| 96 |
"single_word": false,
|
| 97 |
-
"special":
|
| 98 |
},
|
| 99 |
"57529": {
|
| 100 |
-
"content": "
|
| 101 |
"lstrip": false,
|
| 102 |
"normalized": true,
|
| 103 |
"rstrip": false,
|
|
@@ -105,7 +105,15 @@
|
|
| 105 |
"special": false
|
| 106 |
},
|
| 107 |
"57530": {
|
| 108 |
-
"content": "<
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 109 |
"lstrip": false,
|
| 110 |
"normalized": true,
|
| 111 |
"rstrip": false,
|
|
@@ -114,8 +122,15 @@
|
|
| 114 |
}
|
| 115 |
},
|
| 116 |
"additional_special_tokens": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
"<s_iitcdip>",
|
| 118 |
-
"<
|
|
|
|
|
|
|
| 119 |
],
|
| 120 |
"bos_token": "<s>",
|
| 121 |
"clean_up_tokenization_spaces": true,
|
|
|
|
| 49 |
"special": false
|
| 50 |
},
|
| 51 |
"57523": {
|
| 52 |
+
"content": "</s_answer>",
|
| 53 |
"lstrip": false,
|
| 54 |
+
"normalized": true,
|
| 55 |
"rstrip": false,
|
| 56 |
"single_word": false,
|
| 57 |
+
"special": false
|
| 58 |
},
|
| 59 |
"57524": {
|
| 60 |
+
"content": "</s_question>",
|
| 61 |
"lstrip": false,
|
| 62 |
+
"normalized": true,
|
| 63 |
"rstrip": false,
|
| 64 |
"single_word": false,
|
| 65 |
+
"special": false
|
| 66 |
},
|
| 67 |
"57525": {
|
| 68 |
+
"content": "<no/>",
|
| 69 |
"lstrip": false,
|
| 70 |
"normalized": true,
|
| 71 |
"rstrip": false,
|
|
|
|
| 73 |
"special": false
|
| 74 |
},
|
| 75 |
"57526": {
|
| 76 |
+
"content": "<s_answer>",
|
| 77 |
"lstrip": false,
|
| 78 |
"normalized": true,
|
| 79 |
"rstrip": false,
|
|
|
|
| 81 |
"special": false
|
| 82 |
},
|
| 83 |
"57527": {
|
| 84 |
+
"content": "<s_docvqa>",
|
| 85 |
"lstrip": false,
|
| 86 |
"normalized": true,
|
| 87 |
"rstrip": false,
|
|
|
|
| 89 |
"special": false
|
| 90 |
},
|
| 91 |
"57528": {
|
| 92 |
+
"content": "<s_iitcdip>",
|
| 93 |
"lstrip": false,
|
| 94 |
+
"normalized": false,
|
| 95 |
"rstrip": false,
|
| 96 |
"single_word": false,
|
| 97 |
+
"special": true
|
| 98 |
},
|
| 99 |
"57529": {
|
| 100 |
+
"content": "<s_question>",
|
| 101 |
"lstrip": false,
|
| 102 |
"normalized": true,
|
| 103 |
"rstrip": false,
|
|
|
|
| 105 |
"special": false
|
| 106 |
},
|
| 107 |
"57530": {
|
| 108 |
+
"content": "<s_synthdog>",
|
| 109 |
+
"lstrip": false,
|
| 110 |
+
"normalized": false,
|
| 111 |
+
"rstrip": false,
|
| 112 |
+
"single_word": false,
|
| 113 |
+
"special": true
|
| 114 |
+
},
|
| 115 |
+
"57531": {
|
| 116 |
+
"content": "<yes/>",
|
| 117 |
"lstrip": false,
|
| 118 |
"normalized": true,
|
| 119 |
"rstrip": false,
|
|
|
|
| 122 |
}
|
| 123 |
},
|
| 124 |
"additional_special_tokens": [
|
| 125 |
+
"</s_answer>",
|
| 126 |
+
"</s_question>",
|
| 127 |
+
"<no/>",
|
| 128 |
+
"<s_answer>",
|
| 129 |
+
"<s_docvqa>",
|
| 130 |
"<s_iitcdip>",
|
| 131 |
+
"<s_question>",
|
| 132 |
+
"<s_synthdog>",
|
| 133 |
+
"<yes/>"
|
| 134 |
],
|
| 135 |
"bos_token": "<s>",
|
| 136 |
"clean_up_tokenization_spaces": true,
|