Commit
·
b8b1298
1
Parent(s):
4e22580
Upload tiny models for MarkupLMModel
Browse files- added_tokens.json +4 -0
- config.json +32 -0
- generation_config.json +7 -0
- merges.txt +0 -0
- preprocessor_config.json +4 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +287 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<end-of-node>": 50266,
|
| 3 |
+
"[empty-title]": 50265
|
| 4 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MarkupLMModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 32,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 37,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_depth": 10,
|
| 16 |
+
"max_position_embeddings": 512,
|
| 17 |
+
"max_xpath_subs_unit_embeddings": 30,
|
| 18 |
+
"max_xpath_tag_unit_embeddings": 20,
|
| 19 |
+
"model_type": "markuplm",
|
| 20 |
+
"num_attention_heads": 4,
|
| 21 |
+
"num_hidden_layers": 5,
|
| 22 |
+
"pad_token_id": 1,
|
| 23 |
+
"position_embedding_type": "absolute",
|
| 24 |
+
"subs_pad_id": 2,
|
| 25 |
+
"tag_pad_id": 2,
|
| 26 |
+
"torch_dtype": "float32",
|
| 27 |
+
"transformers_version": "4.28.0.dev0",
|
| 28 |
+
"type_vocab_size": 16,
|
| 29 |
+
"use_cache": true,
|
| 30 |
+
"vocab_size": 50267,
|
| 31 |
+
"xpath_unit_hidden_size": 32
|
| 32 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 1,
|
| 6 |
+
"transformers_version": "4.28.0.dev0"
|
| 7 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"feature_extractor_type": "MarkupLMFeatureExtractor",
|
| 3 |
+
"processor_class": "MarkupLMProcessor"
|
| 4 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:401444ec72903ac1067027c75d4a6f70fdbb2b8cd1a8f9594e4641b36d04be34
|
| 3 |
+
size 6968597
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,287 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": {
|
| 4 |
+
"__type": "AddedToken",
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false
|
| 10 |
+
},
|
| 11 |
+
"clean_up_tokenization_spaces": true,
|
| 12 |
+
"cls_token": {
|
| 13 |
+
"__type": "AddedToken",
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": true,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
},
|
| 20 |
+
"eos_token": {
|
| 21 |
+
"__type": "AddedToken",
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": true,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false
|
| 27 |
+
},
|
| 28 |
+
"errors": "replace",
|
| 29 |
+
"mask_token": {
|
| 30 |
+
"__type": "AddedToken",
|
| 31 |
+
"content": "<mask>",
|
| 32 |
+
"lstrip": true,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"max_depth": 50,
|
| 38 |
+
"max_width": 1000,
|
| 39 |
+
"model_max_length": 512,
|
| 40 |
+
"only_label_first_subword": true,
|
| 41 |
+
"pad_token": {
|
| 42 |
+
"__type": "AddedToken",
|
| 43 |
+
"content": "<pad>",
|
| 44 |
+
"lstrip": false,
|
| 45 |
+
"normalized": true,
|
| 46 |
+
"rstrip": false,
|
| 47 |
+
"single_word": false
|
| 48 |
+
},
|
| 49 |
+
"pad_token_label": -100,
|
| 50 |
+
"pad_width": 1001,
|
| 51 |
+
"sep_token": {
|
| 52 |
+
"__type": "AddedToken",
|
| 53 |
+
"content": "</s>",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": true,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false
|
| 58 |
+
},
|
| 59 |
+
"special_tokens_map_file": null,
|
| 60 |
+
"tags_dict": {
|
| 61 |
+
"a": 0,
|
| 62 |
+
"abbr": 1,
|
| 63 |
+
"acronym": 2,
|
| 64 |
+
"address": 3,
|
| 65 |
+
"altGlyph": 4,
|
| 66 |
+
"altGlyphDef": 5,
|
| 67 |
+
"altGlyphItem": 6,
|
| 68 |
+
"animate": 7,
|
| 69 |
+
"animateColor": 8,
|
| 70 |
+
"animateMotion": 9,
|
| 71 |
+
"animateTransform": 10,
|
| 72 |
+
"applet": 11,
|
| 73 |
+
"area": 12,
|
| 74 |
+
"article": 13,
|
| 75 |
+
"aside": 14,
|
| 76 |
+
"audio": 15,
|
| 77 |
+
"b": 16,
|
| 78 |
+
"base": 17,
|
| 79 |
+
"basefont": 18,
|
| 80 |
+
"bdi": 19,
|
| 81 |
+
"bdo": 20,
|
| 82 |
+
"bgsound": 21,
|
| 83 |
+
"big": 22,
|
| 84 |
+
"blink": 23,
|
| 85 |
+
"blockquote": 24,
|
| 86 |
+
"body": 25,
|
| 87 |
+
"br": 26,
|
| 88 |
+
"button": 27,
|
| 89 |
+
"canvas": 28,
|
| 90 |
+
"caption": 29,
|
| 91 |
+
"center": 30,
|
| 92 |
+
"circle": 31,
|
| 93 |
+
"cite": 32,
|
| 94 |
+
"clipPath": 33,
|
| 95 |
+
"code": 34,
|
| 96 |
+
"col": 35,
|
| 97 |
+
"colgroup": 36,
|
| 98 |
+
"color-profile": 37,
|
| 99 |
+
"content": 38,
|
| 100 |
+
"cursor": 39,
|
| 101 |
+
"data": 40,
|
| 102 |
+
"datalist": 41,
|
| 103 |
+
"dd": 42,
|
| 104 |
+
"defs": 43,
|
| 105 |
+
"del": 44,
|
| 106 |
+
"desc": 45,
|
| 107 |
+
"details": 46,
|
| 108 |
+
"dfn": 47,
|
| 109 |
+
"dialog": 48,
|
| 110 |
+
"dir": 49,
|
| 111 |
+
"div": 50,
|
| 112 |
+
"dl": 51,
|
| 113 |
+
"dt": 52,
|
| 114 |
+
"ellipse": 53,
|
| 115 |
+
"em": 54,
|
| 116 |
+
"embed": 55,
|
| 117 |
+
"feBlend": 56,
|
| 118 |
+
"feColorMatrix": 57,
|
| 119 |
+
"feComponentTransfer": 58,
|
| 120 |
+
"feComposite": 59,
|
| 121 |
+
"feConvolveMatrix": 60,
|
| 122 |
+
"feDiffuseLighting": 61,
|
| 123 |
+
"feDisplacementMap": 62,
|
| 124 |
+
"feDistantLight": 63,
|
| 125 |
+
"feFlood": 64,
|
| 126 |
+
"feFuncA": 65,
|
| 127 |
+
"feFuncB": 66,
|
| 128 |
+
"feFuncG": 67,
|
| 129 |
+
"feFuncR": 68,
|
| 130 |
+
"feGaussianBlur": 69,
|
| 131 |
+
"feImage": 70,
|
| 132 |
+
"feMerge": 71,
|
| 133 |
+
"feMergeNode": 72,
|
| 134 |
+
"feMorphology": 73,
|
| 135 |
+
"feOffset": 74,
|
| 136 |
+
"fePointLight": 75,
|
| 137 |
+
"feSpecularLighting": 76,
|
| 138 |
+
"feSpotLight": 77,
|
| 139 |
+
"feTile": 78,
|
| 140 |
+
"feTurbulence": 79,
|
| 141 |
+
"fieldset": 80,
|
| 142 |
+
"figcaption": 81,
|
| 143 |
+
"figure": 82,
|
| 144 |
+
"filter": 83,
|
| 145 |
+
"font": 89,
|
| 146 |
+
"font-face": 88,
|
| 147 |
+
"font-face-format": 84,
|
| 148 |
+
"font-face-name": 85,
|
| 149 |
+
"font-face-src": 86,
|
| 150 |
+
"font-face-uri": 87,
|
| 151 |
+
"footer": 90,
|
| 152 |
+
"foreignObject": 91,
|
| 153 |
+
"form": 92,
|
| 154 |
+
"frame": 93,
|
| 155 |
+
"frameset": 94,
|
| 156 |
+
"g": 95,
|
| 157 |
+
"glyph": 96,
|
| 158 |
+
"glyphRef": 97,
|
| 159 |
+
"h1": 98,
|
| 160 |
+
"h2": 99,
|
| 161 |
+
"h3": 100,
|
| 162 |
+
"h4": 101,
|
| 163 |
+
"h5": 102,
|
| 164 |
+
"h6": 103,
|
| 165 |
+
"head": 104,
|
| 166 |
+
"header": 105,
|
| 167 |
+
"hgroup": 106,
|
| 168 |
+
"hkern": 107,
|
| 169 |
+
"hr": 108,
|
| 170 |
+
"html": 109,
|
| 171 |
+
"i": 110,
|
| 172 |
+
"iframe": 111,
|
| 173 |
+
"image": 112,
|
| 174 |
+
"img": 113,
|
| 175 |
+
"input": 114,
|
| 176 |
+
"ins": 115,
|
| 177 |
+
"kbd": 116,
|
| 178 |
+
"keygen": 117,
|
| 179 |
+
"label": 118,
|
| 180 |
+
"legend": 119,
|
| 181 |
+
"li": 120,
|
| 182 |
+
"line": 121,
|
| 183 |
+
"linearGradient": 122,
|
| 184 |
+
"link": 123,
|
| 185 |
+
"main": 124,
|
| 186 |
+
"map": 125,
|
| 187 |
+
"mark": 126,
|
| 188 |
+
"marker": 127,
|
| 189 |
+
"marquee": 128,
|
| 190 |
+
"mask": 129,
|
| 191 |
+
"math": 130,
|
| 192 |
+
"menu": 131,
|
| 193 |
+
"menuitem": 132,
|
| 194 |
+
"meta": 133,
|
| 195 |
+
"metadata": 134,
|
| 196 |
+
"meter": 135,
|
| 197 |
+
"missing-glyph": 136,
|
| 198 |
+
"mpath": 137,
|
| 199 |
+
"nav": 138,
|
| 200 |
+
"nobr": 139,
|
| 201 |
+
"noembed": 140,
|
| 202 |
+
"noframes": 141,
|
| 203 |
+
"noscript": 142,
|
| 204 |
+
"object": 143,
|
| 205 |
+
"ol": 144,
|
| 206 |
+
"optgroup": 145,
|
| 207 |
+
"option": 146,
|
| 208 |
+
"output": 147,
|
| 209 |
+
"p": 148,
|
| 210 |
+
"param": 149,
|
| 211 |
+
"path": 150,
|
| 212 |
+
"pattern": 151,
|
| 213 |
+
"picture": 152,
|
| 214 |
+
"plaintext": 153,
|
| 215 |
+
"polygon": 154,
|
| 216 |
+
"polyline": 155,
|
| 217 |
+
"portal": 156,
|
| 218 |
+
"pre": 157,
|
| 219 |
+
"progress": 158,
|
| 220 |
+
"q": 159,
|
| 221 |
+
"radialGradient": 160,
|
| 222 |
+
"rb": 161,
|
| 223 |
+
"rect": 162,
|
| 224 |
+
"rp": 163,
|
| 225 |
+
"rt": 164,
|
| 226 |
+
"rtc": 165,
|
| 227 |
+
"ruby": 166,
|
| 228 |
+
"s": 167,
|
| 229 |
+
"samp": 168,
|
| 230 |
+
"script": 169,
|
| 231 |
+
"section": 170,
|
| 232 |
+
"select": 171,
|
| 233 |
+
"set": 172,
|
| 234 |
+
"shadow": 173,
|
| 235 |
+
"slot": 174,
|
| 236 |
+
"small": 175,
|
| 237 |
+
"source": 176,
|
| 238 |
+
"spacer": 177,
|
| 239 |
+
"span": 178,
|
| 240 |
+
"stop": 179,
|
| 241 |
+
"strike": 180,
|
| 242 |
+
"strong": 181,
|
| 243 |
+
"style": 182,
|
| 244 |
+
"sub": 183,
|
| 245 |
+
"summary": 184,
|
| 246 |
+
"sup": 185,
|
| 247 |
+
"svg": 186,
|
| 248 |
+
"switch": 187,
|
| 249 |
+
"symbol": 188,
|
| 250 |
+
"table": 189,
|
| 251 |
+
"tbody": 190,
|
| 252 |
+
"td": 191,
|
| 253 |
+
"template": 192,
|
| 254 |
+
"text": 193,
|
| 255 |
+
"textPath": 194,
|
| 256 |
+
"textarea": 195,
|
| 257 |
+
"tfoot": 196,
|
| 258 |
+
"th": 197,
|
| 259 |
+
"thead": 198,
|
| 260 |
+
"time": 199,
|
| 261 |
+
"title": 200,
|
| 262 |
+
"tr": 201,
|
| 263 |
+
"track": 202,
|
| 264 |
+
"tref": 203,
|
| 265 |
+
"tspan": 204,
|
| 266 |
+
"tt": 205,
|
| 267 |
+
"u": 206,
|
| 268 |
+
"ul": 207,
|
| 269 |
+
"use": 208,
|
| 270 |
+
"var": 209,
|
| 271 |
+
"video": 210,
|
| 272 |
+
"view": 211,
|
| 273 |
+
"vkern": 212,
|
| 274 |
+
"wbr": 213,
|
| 275 |
+
"xmp": 214
|
| 276 |
+
},
|
| 277 |
+
"tokenizer_class": "MarkupLMTokenizer",
|
| 278 |
+
"trim_offsets": false,
|
| 279 |
+
"unk_token": {
|
| 280 |
+
"__type": "AddedToken",
|
| 281 |
+
"content": "<unk>",
|
| 282 |
+
"lstrip": false,
|
| 283 |
+
"normalized": true,
|
| 284 |
+
"rstrip": false,
|
| 285 |
+
"single_word": false
|
| 286 |
+
}
|
| 287 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|