hilariooliveira commited on
Commit
b510437
1 Parent(s): a86b069

End of training

Browse files
README.md ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: dandelin/vilt-b32-mlm
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - vqa
8
+ model-index:
9
+ - name: vilt_finetuned_200
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_finetuned_200
17
+
18
+ This model is a fine-tuned version of [dandelin/vilt-b32-mlm](https://huggingface.co/dandelin/vilt-b32-mlm) on the vqa 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: 2
44
+
45
+ ### Training results
46
+
47
+
48
+
49
+ ### Framework versions
50
+
51
+ - Transformers 4.40.2
52
+ - Pytorch 2.2.1+cu121
53
+ - Datasets 2.19.1
54
+ - Tokenizers 0.19.1
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": "french",
12
+ "1": "plain",
13
+ "2": "arrow",
14
+ "3": "green",
15
+ "4": "skateboard",
16
+ "5": "down",
17
+ "6": "protection",
18
+ "7": "lady",
19
+ "8": "snowboard",
20
+ "9": "black",
21
+ "10": "lanyard",
22
+ "11": "cross",
23
+ "12": "style",
24
+ "13": "man",
25
+ "14": "don't know",
26
+ "15": "africa",
27
+ "16": "4",
28
+ "17": "hawaii",
29
+ "18": "boy",
30
+ "19": "sidewalk",
31
+ "20": "skier",
32
+ "21": "walking",
33
+ "22": "8:35",
34
+ "23": "king",
35
+ "24": "door",
36
+ "25": "bricks",
37
+ "26": "security",
38
+ "27": "church",
39
+ "28": "window",
40
+ "29": "gray",
41
+ "30": "skiing",
42
+ "31": "white",
43
+ "32": "red and yellow",
44
+ "33": "tv",
45
+ "34": "canopy",
46
+ "35": "cup",
47
+ "36": "giraffes",
48
+ "37": "red and blue",
49
+ "38": "camera",
50
+ "39": "shrimp",
51
+ "40": "blue",
52
+ "41": "suv",
53
+ "42": "crossing",
54
+ "43": "in car",
55
+ "44": "roof",
56
+ "45": "talking",
57
+ "46": "queen",
58
+ "47": "on road",
59
+ "48": "shelter",
60
+ "49": "women",
61
+ "50": "tan",
62
+ "51": "snow",
63
+ "52": "smiling",
64
+ "53": "bike rack",
65
+ "54": "black and white",
66
+ "55": "fence",
67
+ "56": "smile",
68
+ "57": "5",
69
+ "58": "sleeping",
70
+ "59": "table",
71
+ "60": "exit",
72
+ "61": "picnic table",
73
+ "62": "dog",
74
+ "63": "windows",
75
+ "64": "right",
76
+ "65": "purple",
77
+ "66": "many",
78
+ "67": "blonde",
79
+ "68": "woman",
80
+ "69": "calico",
81
+ "70": "7",
82
+ "71": "monitor",
83
+ "72": "0",
84
+ "73": "low",
85
+ "74": "zoo",
86
+ "75": "tired",
87
+ "76": "street",
88
+ "77": "person",
89
+ "78": "yes",
90
+ "79": "pink",
91
+ "80": "double",
92
+ "81": "clear",
93
+ "82": "gray and black",
94
+ "83": "natural",
95
+ "84": "ground",
96
+ "85": "lg",
97
+ "86": "can't tell",
98
+ "87": "clock tower",
99
+ "88": "shadow",
100
+ "89": "red",
101
+ "90": "blue and white",
102
+ "91": "soccer",
103
+ "92": "girl",
104
+ "93": "bikes",
105
+ "94": "air",
106
+ "95": "beagle",
107
+ "96": "soccer ball",
108
+ "97": "nothing",
109
+ "98": "screen",
110
+ "99": "stand",
111
+ "100": "not sure",
112
+ "101": "orange",
113
+ "102": "snowboarding",
114
+ "103": "2013",
115
+ "104": "shadows",
116
+ "105": "laying down",
117
+ "106": "wine tasting",
118
+ "107": "photographer",
119
+ "108": "happy",
120
+ "109": "yellow",
121
+ "110": "bedroom",
122
+ "111": "solid",
123
+ "112": "wall",
124
+ "113": "station",
125
+ "114": "ball",
126
+ "115": "hair",
127
+ "116": "sky",
128
+ "117": "crown",
129
+ "118": "2010",
130
+ "119": "not there",
131
+ "120": "lying down",
132
+ "121": "plastic",
133
+ "122": "large",
134
+ "123": "white and black",
135
+ "124": "cloudy",
136
+ "125": "6",
137
+ "126": "platform",
138
+ "127": "they aren't",
139
+ "128": "curtains",
140
+ "129": "birthday",
141
+ "130": "hat",
142
+ "131": "neon",
143
+ "132": "snowboarder",
144
+ "133": "tower",
145
+ "134": "watching",
146
+ "135": "full",
147
+ "136": "backpack",
148
+ "137": "doughnut",
149
+ "138": "cage",
150
+ "139": "sun",
151
+ "140": "1",
152
+ "141": "resting",
153
+ "142": "trees",
154
+ "143": "donut",
155
+ "144": "train",
156
+ "145": "talking on phone",
157
+ "146": "bicycles",
158
+ "147": "beige",
159
+ "148": "9:35",
160
+ "149": "jeep",
161
+ "150": "clock",
162
+ "151": "name tag",
163
+ "152": "ice cream",
164
+ "153": "white and blue",
165
+ "154": "bus",
166
+ "155": "3",
167
+ "156": "brown",
168
+ "157": "outside",
169
+ "158": "giraffe",
170
+ "159": "out",
171
+ "160": "chair",
172
+ "161": "big ben",
173
+ "162": "small",
174
+ "163": "bicycle",
175
+ "164": "2000",
176
+ "165": "woods",
177
+ "166": "unknown",
178
+ "167": "shade",
179
+ "168": "net",
180
+ "169": "desert",
181
+ "170": "little girl",
182
+ "171": "fashion",
183
+ "172": "cat",
184
+ "173": "2",
185
+ "174": "tabby",
186
+ "175": "leather",
187
+ "176": "7:45",
188
+ "177": "chopsticks",
189
+ "178": "on street",
190
+ "179": "no",
191
+ "180": "rack",
192
+ "181": "park",
193
+ "182": "stripes",
194
+ "183": "skateboarding",
195
+ "184": "dirt",
196
+ "185": "car",
197
+ "186": "wedding",
198
+ "187": "necklace",
199
+ "188": "curtain",
200
+ "189": "10",
201
+ "190": "at table",
202
+ "191": "forest",
203
+ "192": "plate",
204
+ "193": "human",
205
+ "194": "brick",
206
+ "195": "tent",
207
+ "196": "8",
208
+ "197": "7:35",
209
+ "198": "wine"
210
+ },
211
+ "image_size": 384,
212
+ "initializer_range": 0.02,
213
+ "intermediate_size": 3072,
214
+ "label2id": {
215
+ "0": 72,
216
+ "1": 140,
217
+ "10": 189,
218
+ "2": 173,
219
+ "2000": 164,
220
+ "2010": 118,
221
+ "2013": 103,
222
+ "3": 155,
223
+ "4": 16,
224
+ "5": 57,
225
+ "6": 125,
226
+ "7": 70,
227
+ "7:35": 197,
228
+ "7:45": 176,
229
+ "8": 196,
230
+ "8:35": 22,
231
+ "9:35": 148,
232
+ "africa": 15,
233
+ "air": 94,
234
+ "arrow": 2,
235
+ "at table": 190,
236
+ "backpack": 136,
237
+ "ball": 114,
238
+ "beagle": 95,
239
+ "bedroom": 110,
240
+ "beige": 147,
241
+ "bicycle": 163,
242
+ "bicycles": 146,
243
+ "big ben": 161,
244
+ "bike rack": 53,
245
+ "bikes": 93,
246
+ "birthday": 129,
247
+ "black": 9,
248
+ "black and white": 54,
249
+ "blonde": 67,
250
+ "blue": 40,
251
+ "blue and white": 90,
252
+ "boy": 18,
253
+ "brick": 194,
254
+ "bricks": 25,
255
+ "brown": 156,
256
+ "bus": 154,
257
+ "cage": 138,
258
+ "calico": 69,
259
+ "camera": 38,
260
+ "can't tell": 86,
261
+ "canopy": 34,
262
+ "car": 185,
263
+ "cat": 172,
264
+ "chair": 160,
265
+ "chopsticks": 177,
266
+ "church": 27,
267
+ "clear": 81,
268
+ "clock": 150,
269
+ "clock tower": 87,
270
+ "cloudy": 124,
271
+ "cross": 11,
272
+ "crossing": 42,
273
+ "crown": 117,
274
+ "cup": 35,
275
+ "curtain": 188,
276
+ "curtains": 128,
277
+ "desert": 169,
278
+ "dirt": 184,
279
+ "dog": 62,
280
+ "don't know": 14,
281
+ "donut": 143,
282
+ "door": 24,
283
+ "double": 80,
284
+ "doughnut": 137,
285
+ "down": 5,
286
+ "exit": 60,
287
+ "fashion": 171,
288
+ "fence": 55,
289
+ "forest": 191,
290
+ "french": 0,
291
+ "full": 135,
292
+ "giraffe": 158,
293
+ "giraffes": 36,
294
+ "girl": 92,
295
+ "gray": 29,
296
+ "gray and black": 82,
297
+ "green": 3,
298
+ "ground": 84,
299
+ "hair": 115,
300
+ "happy": 108,
301
+ "hat": 130,
302
+ "hawaii": 17,
303
+ "human": 193,
304
+ "ice cream": 152,
305
+ "in car": 43,
306
+ "jeep": 149,
307
+ "king": 23,
308
+ "lady": 7,
309
+ "lanyard": 10,
310
+ "large": 122,
311
+ "laying down": 105,
312
+ "leather": 175,
313
+ "lg": 85,
314
+ "little girl": 170,
315
+ "low": 73,
316
+ "lying down": 120,
317
+ "man": 13,
318
+ "many": 66,
319
+ "monitor": 71,
320
+ "name tag": 151,
321
+ "natural": 83,
322
+ "necklace": 187,
323
+ "neon": 131,
324
+ "net": 168,
325
+ "no": 179,
326
+ "not sure": 100,
327
+ "not there": 119,
328
+ "nothing": 97,
329
+ "on road": 47,
330
+ "on street": 178,
331
+ "orange": 101,
332
+ "out": 159,
333
+ "outside": 157,
334
+ "park": 181,
335
+ "person": 77,
336
+ "photographer": 107,
337
+ "picnic table": 61,
338
+ "pink": 79,
339
+ "plain": 1,
340
+ "plastic": 121,
341
+ "plate": 192,
342
+ "platform": 126,
343
+ "protection": 6,
344
+ "purple": 65,
345
+ "queen": 46,
346
+ "rack": 180,
347
+ "red": 89,
348
+ "red and blue": 37,
349
+ "red and yellow": 32,
350
+ "resting": 141,
351
+ "right": 64,
352
+ "roof": 44,
353
+ "screen": 98,
354
+ "security": 26,
355
+ "shade": 167,
356
+ "shadow": 88,
357
+ "shadows": 104,
358
+ "shelter": 48,
359
+ "shrimp": 39,
360
+ "sidewalk": 19,
361
+ "skateboard": 4,
362
+ "skateboarding": 183,
363
+ "skier": 20,
364
+ "skiing": 30,
365
+ "sky": 116,
366
+ "sleeping": 58,
367
+ "small": 162,
368
+ "smile": 56,
369
+ "smiling": 52,
370
+ "snow": 51,
371
+ "snowboard": 8,
372
+ "snowboarder": 132,
373
+ "snowboarding": 102,
374
+ "soccer": 91,
375
+ "soccer ball": 96,
376
+ "solid": 111,
377
+ "stand": 99,
378
+ "station": 113,
379
+ "street": 76,
380
+ "stripes": 182,
381
+ "style": 12,
382
+ "sun": 139,
383
+ "suv": 41,
384
+ "tabby": 174,
385
+ "table": 59,
386
+ "talking": 45,
387
+ "talking on phone": 145,
388
+ "tan": 50,
389
+ "tent": 195,
390
+ "they aren't": 127,
391
+ "tired": 75,
392
+ "tower": 133,
393
+ "train": 144,
394
+ "trees": 142,
395
+ "tv": 33,
396
+ "unknown": 166,
397
+ "walking": 21,
398
+ "wall": 112,
399
+ "watching": 134,
400
+ "wedding": 186,
401
+ "white": 31,
402
+ "white and black": 123,
403
+ "white and blue": 153,
404
+ "window": 28,
405
+ "windows": 63,
406
+ "wine": 198,
407
+ "wine tasting": 106,
408
+ "woman": 68,
409
+ "women": 49,
410
+ "woods": 165,
411
+ "yellow": 109,
412
+ "yes": 78,
413
+ "zoo": 74
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": "4.40.2",
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:cff1a3615424760bc37e5f92ea80ba7ecf76fd9d50d6ee472919feca2c393e34
3
+ size 452365324
preprocessor_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_valid_processor_keys": [
3
+ "images",
4
+ "do_resize",
5
+ "size",
6
+ "size_divisor",
7
+ "resample",
8
+ "do_rescale",
9
+ "rescale_factor",
10
+ "do_normalize",
11
+ "image_mean",
12
+ "image_std",
13
+ "do_pad",
14
+ "return_tensors",
15
+ "data_format",
16
+ "input_data_format"
17
+ ],
18
+ "do_normalize": true,
19
+ "do_pad": true,
20
+ "do_rescale": true,
21
+ "do_resize": true,
22
+ "image_mean": [
23
+ 0.5,
24
+ 0.5,
25
+ 0.5
26
+ ],
27
+ "image_processor_type": "ViltImageProcessor",
28
+ "image_std": [
29
+ 0.5,
30
+ 0.5,
31
+ 0.5
32
+ ],
33
+ "processor_class": "ViltProcessor",
34
+ "resample": 3,
35
+ "rescale_factor": 0.00392156862745098,
36
+ "size": {
37
+ "shortest_edge": 384
38
+ },
39
+ "size_divisor": 32
40
+ }
runs/May14_23-57-52_8a267db08b19/events.out.tfevents.1715731073.8a267db08b19.4972.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe2b5a0298820697354d47d1533bebbf2274748540800bd2d270079ee681ad6a
3
+ size 12575
runs/May15_00-04-57_8a267db08b19/events.out.tfevents.1715731503.8a267db08b19.4972.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:913bc22ecf021d00e5aadf9a9160afd1ce8c20ad9fc7fc7601b076a650667338
3
+ size 13337
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 sha256:4d4c10000404adc0627ac2ef0c1424e640a163a9f11777305836458b2b500944
3
+ size 4984
vocab.txt ADDED
The diff for this file is too large to render. See raw diff