HelloWorld2307 commited on
Commit
0824c04
1 Parent(s): bc90912

End of training

Browse files
README.md CHANGED
@@ -1,3 +1,54 @@
1
  ---
2
  license: apache-2.0
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
+ base_model: dandelin/vilt-b32-mlm
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - vqa_v2
8
+ model-index:
9
+ - name: ViLT
10
+ results: []
11
  ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # ViLT
17
+
18
+ This model is a fine-tuned version of [dandelin/vilt-b32-mlm](https://huggingface.co/dandelin/vilt-b32-mlm) on the vqa_v2 dataset.
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 5e-05
38
+ - train_batch_size: 4
39
+ - eval_batch_size: 8
40
+ - seed: 42
41
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
42
+ - lr_scheduler_type: linear
43
+ - num_epochs: 20
44
+
45
+ ### Training results
46
+
47
+
48
+
49
+ ### Framework versions
50
+
51
+ - Transformers 4.37.0
52
+ - Pytorch 2.1.2
53
+ - Datasets 2.1.0
54
+ - Tokenizers 0.15.1
config.json ADDED
@@ -0,0 +1,431 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "dandelin/vilt-b32-mlm",
3
+ "architectures": [
4
+ "ViltForQuestionAnswering"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.0,
9
+ "hidden_size": 768,
10
+ "id2label": {
11
+ "0": "king",
12
+ "1": "4",
13
+ "2": "canopy",
14
+ "3": "sun",
15
+ "4": "calico",
16
+ "5": "brown",
17
+ "6": "style",
18
+ "7": "cloudy",
19
+ "8": "blue",
20
+ "9": "birthday",
21
+ "10": "bicycles",
22
+ "11": "can't tell",
23
+ "12": "forest",
24
+ "13": "doughnut",
25
+ "14": "neon",
26
+ "15": "picnic table",
27
+ "16": "watching",
28
+ "17": "wall",
29
+ "18": "fence",
30
+ "19": "chair",
31
+ "20": "down",
32
+ "21": "stripes",
33
+ "22": "orange",
34
+ "23": "hat",
35
+ "24": "cat",
36
+ "25": "skateboarding",
37
+ "26": "exit",
38
+ "27": "shade",
39
+ "28": "sleeping",
40
+ "29": "backpack",
41
+ "30": "brick",
42
+ "31": "net",
43
+ "32": "platform",
44
+ "33": "laying down",
45
+ "34": "boy",
46
+ "35": "hair",
47
+ "36": "camera",
48
+ "37": "smile",
49
+ "38": "name tag",
50
+ "39": "in car",
51
+ "40": "bedroom",
52
+ "41": "snowboarder",
53
+ "42": "not there",
54
+ "43": "sky",
55
+ "44": "5",
56
+ "45": "chopsticks",
57
+ "46": "trees",
58
+ "47": "table",
59
+ "48": "6",
60
+ "49": "natural",
61
+ "50": "door",
62
+ "51": "africa",
63
+ "52": "smiling",
64
+ "53": "10",
65
+ "54": "rack",
66
+ "55": "7",
67
+ "56": "fashion",
68
+ "57": "snowboarding",
69
+ "58": "skier",
70
+ "59": "shelter",
71
+ "60": "photographer",
72
+ "61": "2",
73
+ "62": "happy",
74
+ "63": "resting",
75
+ "64": "on street",
76
+ "65": "7:35",
77
+ "66": "yellow",
78
+ "67": "air",
79
+ "68": "beagle",
80
+ "69": "plate",
81
+ "70": "protection",
82
+ "71": "not sure",
83
+ "72": "human",
84
+ "73": "9:35",
85
+ "74": "bikes",
86
+ "75": "cage",
87
+ "76": "hawaii",
88
+ "77": "out",
89
+ "78": "on road",
90
+ "79": "2013",
91
+ "80": "2010",
92
+ "81": "skiing",
93
+ "82": "station",
94
+ "83": "ice cream",
95
+ "84": "window",
96
+ "85": "tent",
97
+ "86": "clear",
98
+ "87": "red",
99
+ "88": "stand",
100
+ "89": "wine tasting",
101
+ "90": "soccer",
102
+ "91": "lying down",
103
+ "92": "queen",
104
+ "93": "crown",
105
+ "94": "shrimp",
106
+ "95": "woods",
107
+ "96": "black and white",
108
+ "97": "tired",
109
+ "98": "ground",
110
+ "99": "3",
111
+ "100": "suv",
112
+ "101": "clock tower",
113
+ "102": "clock",
114
+ "103": "gray and black",
115
+ "104": "bike rack",
116
+ "105": "green",
117
+ "106": "necklace",
118
+ "107": "plastic",
119
+ "108": "girl",
120
+ "109": "street",
121
+ "110": "full",
122
+ "111": "desert",
123
+ "112": "security",
124
+ "113": "no",
125
+ "114": "yes",
126
+ "115": "lg",
127
+ "116": "curtains",
128
+ "117": "person",
129
+ "118": "snowboard",
130
+ "119": "bricks",
131
+ "120": "tabby",
132
+ "121": "car",
133
+ "122": "ball",
134
+ "123": "outside",
135
+ "124": "pink",
136
+ "125": "donut",
137
+ "126": "8:35",
138
+ "127": "tower",
139
+ "128": "screen",
140
+ "129": "gray",
141
+ "130": "lanyard",
142
+ "131": "solid",
143
+ "132": "blue and white",
144
+ "133": "women",
145
+ "134": "low",
146
+ "135": "cross",
147
+ "136": "talking on phone",
148
+ "137": "bus",
149
+ "138": "red and blue",
150
+ "139": "wedding",
151
+ "140": "red and yellow",
152
+ "141": "purple",
153
+ "142": "leather",
154
+ "143": "white and black",
155
+ "144": "right",
156
+ "145": "beige",
157
+ "146": "unknown",
158
+ "147": "wine",
159
+ "148": "black",
160
+ "149": "little girl",
161
+ "150": "shadow",
162
+ "151": "train",
163
+ "152": "0",
164
+ "153": "roof",
165
+ "154": "small",
166
+ "155": "large",
167
+ "156": "snow",
168
+ "157": "bicycle",
169
+ "158": "park",
170
+ "159": "1",
171
+ "160": "man",
172
+ "161": "cup",
173
+ "162": "dirt",
174
+ "163": "arrow",
175
+ "164": "big ben",
176
+ "165": "monitor",
177
+ "166": "sidewalk",
178
+ "167": "talking",
179
+ "168": "at table",
180
+ "169": "skateboard",
181
+ "170": "crossing",
182
+ "171": "dog",
183
+ "172": "tan",
184
+ "173": "tv",
185
+ "174": "giraffe",
186
+ "175": "they aren't",
187
+ "176": "woman",
188
+ "177": "church",
189
+ "178": "7:45",
190
+ "179": "8",
191
+ "180": "french",
192
+ "181": "2000",
193
+ "182": "soccer ball",
194
+ "183": "giraffes",
195
+ "184": "double",
196
+ "185": "jeep",
197
+ "186": "white",
198
+ "187": "shadows",
199
+ "188": "zoo",
200
+ "189": "nothing",
201
+ "190": "don't know",
202
+ "191": "walking",
203
+ "192": "blonde",
204
+ "193": "plain",
205
+ "194": "white and blue",
206
+ "195": "curtain",
207
+ "196": "lady",
208
+ "197": "many",
209
+ "198": "windows"
210
+ },
211
+ "image_size": 384,
212
+ "initializer_range": 0.02,
213
+ "intermediate_size": 3072,
214
+ "label2id": {
215
+ "0": 152,
216
+ "1": 159,
217
+ "10": 53,
218
+ "2": 61,
219
+ "2000": 181,
220
+ "2010": 80,
221
+ "2013": 79,
222
+ "3": 99,
223
+ "4": 1,
224
+ "5": 44,
225
+ "6": 48,
226
+ "7": 55,
227
+ "7:35": 65,
228
+ "7:45": 178,
229
+ "8": 179,
230
+ "8:35": 126,
231
+ "9:35": 73,
232
+ "africa": 51,
233
+ "air": 67,
234
+ "arrow": 163,
235
+ "at table": 168,
236
+ "backpack": 29,
237
+ "ball": 122,
238
+ "beagle": 68,
239
+ "bedroom": 40,
240
+ "beige": 145,
241
+ "bicycle": 157,
242
+ "bicycles": 10,
243
+ "big ben": 164,
244
+ "bike rack": 104,
245
+ "bikes": 74,
246
+ "birthday": 9,
247
+ "black": 148,
248
+ "black and white": 96,
249
+ "blonde": 192,
250
+ "blue": 8,
251
+ "blue and white": 132,
252
+ "boy": 34,
253
+ "brick": 30,
254
+ "bricks": 119,
255
+ "brown": 5,
256
+ "bus": 137,
257
+ "cage": 75,
258
+ "calico": 4,
259
+ "camera": 36,
260
+ "can't tell": 11,
261
+ "canopy": 2,
262
+ "car": 121,
263
+ "cat": 24,
264
+ "chair": 19,
265
+ "chopsticks": 45,
266
+ "church": 177,
267
+ "clear": 86,
268
+ "clock": 102,
269
+ "clock tower": 101,
270
+ "cloudy": 7,
271
+ "cross": 135,
272
+ "crossing": 170,
273
+ "crown": 93,
274
+ "cup": 161,
275
+ "curtain": 195,
276
+ "curtains": 116,
277
+ "desert": 111,
278
+ "dirt": 162,
279
+ "dog": 171,
280
+ "don't know": 190,
281
+ "donut": 125,
282
+ "door": 50,
283
+ "double": 184,
284
+ "doughnut": 13,
285
+ "down": 20,
286
+ "exit": 26,
287
+ "fashion": 56,
288
+ "fence": 18,
289
+ "forest": 12,
290
+ "french": 180,
291
+ "full": 110,
292
+ "giraffe": 174,
293
+ "giraffes": 183,
294
+ "girl": 108,
295
+ "gray": 129,
296
+ "gray and black": 103,
297
+ "green": 105,
298
+ "ground": 98,
299
+ "hair": 35,
300
+ "happy": 62,
301
+ "hat": 23,
302
+ "hawaii": 76,
303
+ "human": 72,
304
+ "ice cream": 83,
305
+ "in car": 39,
306
+ "jeep": 185,
307
+ "king": 0,
308
+ "lady": 196,
309
+ "lanyard": 130,
310
+ "large": 155,
311
+ "laying down": 33,
312
+ "leather": 142,
313
+ "lg": 115,
314
+ "little girl": 149,
315
+ "low": 134,
316
+ "lying down": 91,
317
+ "man": 160,
318
+ "many": 197,
319
+ "monitor": 165,
320
+ "name tag": 38,
321
+ "natural": 49,
322
+ "necklace": 106,
323
+ "neon": 14,
324
+ "net": 31,
325
+ "no": 113,
326
+ "not sure": 71,
327
+ "not there": 42,
328
+ "nothing": 189,
329
+ "on road": 78,
330
+ "on street": 64,
331
+ "orange": 22,
332
+ "out": 77,
333
+ "outside": 123,
334
+ "park": 158,
335
+ "person": 117,
336
+ "photographer": 60,
337
+ "picnic table": 15,
338
+ "pink": 124,
339
+ "plain": 193,
340
+ "plastic": 107,
341
+ "plate": 69,
342
+ "platform": 32,
343
+ "protection": 70,
344
+ "purple": 141,
345
+ "queen": 92,
346
+ "rack": 54,
347
+ "red": 87,
348
+ "red and blue": 138,
349
+ "red and yellow": 140,
350
+ "resting": 63,
351
+ "right": 144,
352
+ "roof": 153,
353
+ "screen": 128,
354
+ "security": 112,
355
+ "shade": 27,
356
+ "shadow": 150,
357
+ "shadows": 187,
358
+ "shelter": 59,
359
+ "shrimp": 94,
360
+ "sidewalk": 166,
361
+ "skateboard": 169,
362
+ "skateboarding": 25,
363
+ "skier": 58,
364
+ "skiing": 81,
365
+ "sky": 43,
366
+ "sleeping": 28,
367
+ "small": 154,
368
+ "smile": 37,
369
+ "smiling": 52,
370
+ "snow": 156,
371
+ "snowboard": 118,
372
+ "snowboarder": 41,
373
+ "snowboarding": 57,
374
+ "soccer": 90,
375
+ "soccer ball": 182,
376
+ "solid": 131,
377
+ "stand": 88,
378
+ "station": 82,
379
+ "street": 109,
380
+ "stripes": 21,
381
+ "style": 6,
382
+ "sun": 3,
383
+ "suv": 100,
384
+ "tabby": 120,
385
+ "table": 47,
386
+ "talking": 167,
387
+ "talking on phone": 136,
388
+ "tan": 172,
389
+ "tent": 85,
390
+ "they aren't": 175,
391
+ "tired": 97,
392
+ "tower": 127,
393
+ "train": 151,
394
+ "trees": 46,
395
+ "tv": 173,
396
+ "unknown": 146,
397
+ "walking": 191,
398
+ "wall": 17,
399
+ "watching": 16,
400
+ "wedding": 139,
401
+ "white": 186,
402
+ "white and black": 143,
403
+ "white and blue": 194,
404
+ "window": 84,
405
+ "windows": 198,
406
+ "wine": 147,
407
+ "wine tasting": 89,
408
+ "woman": 176,
409
+ "women": 133,
410
+ "woods": 95,
411
+ "yellow": 66,
412
+ "yes": 114,
413
+ "zoo": 188
414
+ },
415
+ "layer_norm_eps": 1e-12,
416
+ "max_image_length": -1,
417
+ "max_position_embeddings": 40,
418
+ "modality_type_vocab_size": 2,
419
+ "model_type": "vilt",
420
+ "num_attention_heads": 12,
421
+ "num_channels": 3,
422
+ "num_hidden_layers": 12,
423
+ "num_images": -1,
424
+ "patch_size": 32,
425
+ "qkv_bias": true,
426
+ "tie_word_embeddings": false,
427
+ "torch_dtype": "float32",
428
+ "transformers_version": "4.37.0",
429
+ "type_vocab_size": 2,
430
+ "vocab_size": 30522
431
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f06d086bad92222738d30ce8f3c721839a76046f41120907c3f6bbbc200cb8d
3
+ size 452365324
preprocessor_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_pad": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.5,
8
+ 0.5,
9
+ 0.5
10
+ ],
11
+ "image_processor_type": "ViltImageProcessor",
12
+ "image_std": [
13
+ 0.5,
14
+ 0.5,
15
+ 0.5
16
+ ],
17
+ "processor_class": "ViltProcessor",
18
+ "resample": 3,
19
+ "rescale_factor": 0.00392156862745098,
20
+ "size": {
21
+ "shortest_edge": 384
22
+ },
23
+ "size_divisor": 32
24
+ }
runs/Feb24_05-20-17_5d85855224cc/events.out.tfevents.1708752021.5d85855224cc.402.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64e55d6c95a230a672e353a0bc4472fd44233f06d065ec558b7e2170844ee873
3
+ size 15714
special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": true,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "mask_token": "[MASK]",
48
+ "model_max_length": 40,
49
+ "pad_token": "[PAD]",
50
+ "processor_class": "ViltProcessor",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef62d33e4031d9c0e90016a7e242df56fa7187c5b091ee0b1ecb194fe3dd6847
3
+ size 4728
vocab.txt ADDED
The diff for this file is too large to render. See raw diff