Hugging Face's logo

g8a9
/
roberta-tiny-10M

Fill-Mask
Transformers
PyTorch
TensorBoard
roberta
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
g8a9 commited on
Commit
cf8be5e
•
1 Parent(s): 59c8a40

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +96 -130
README.md CHANGED
@@ -15,8 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  This model was trained from scratch on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 6.4867
19
- - Accuracy: 0.1377
20
 
21
  ## Model description
22
 
@@ -35,142 +35,108 @@ More information needed
35
  ### Training hyperparameters
36
 
37
  The following hyperparameters were used during training:
38
- - learning_rate: 2e-05
39
- - train_batch_size: 32
40
- - eval_batch_size: 64
41
  - seed: 42
42
- - gradient_accumulation_steps: 4
43
- - total_train_batch_size: 128
44
  - optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-08
45
- - lr_scheduler_type: constant_with_warmup
46
  - lr_scheduler_warmup_steps: 50
47
- - num_epochs: 40.0
 
48
 
49
  ### Training results
50
 
51
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
52
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
53
- | 10.5287 | 0.26 | 50 | 10.4478 | 0.0481 |
54
- | 10.0073 | 0.51 | 100 | 9.9550 | 0.0488 |
55
- | 9.6268 | 0.77 | 150 | 9.5865 | 0.0488 |
56
- | 9.9837 | 1.03 | 200 | 9.2502 | 0.0471 |
57
- | 8.9701 | 1.29 | 250 | 8.9370 | 0.0466 |
58
- | 8.6689 | 1.54 | 300 | 8.6447 | 0.0473 |
59
- | 8.3893 | 1.8 | 350 | 8.3794 | 0.0473 |
60
- | 8.1697 | 2.06 | 400 | 8.1342 | 0.0506 |
61
- | 7.926 | 2.32 | 450 | 7.9221 | 0.0617 |
62
- | 7.7329 | 2.58 | 500 | 7.7398 | 0.0627 |
63
- | 7.582 | 2.83 | 550 | 7.5844 | 0.0691 |
64
- | 7.4419 | 3.09 | 600 | 7.4620 | 0.0729 |
65
- | 7.3658 | 3.35 | 650 | 7.3735 | 0.0781 |
66
- | 7.2857 | 3.61 | 700 | 7.3049 | 0.0801 |
67
- | 7.224 | 3.86 | 750 | 7.2554 | 0.0831 |
68
- | 7.1851 | 4.12 | 800 | 7.2082 | 0.0853 |
69
- | 7.1327 | 4.38 | 850 | 7.1678 | 0.0878 |
70
- | 7.0947 | 4.64 | 900 | 7.1326 | 0.0909 |
71
- | 7.0761 | 4.89 | 950 | 7.1069 | 0.0919 |
72
- | 7.0551 | 5.15 | 1000 | 7.0806 | 0.0943 |
73
- | 7.0389 | 5.41 | 1050 | 7.0588 | 0.0952 |
74
- | 7.0226 | 5.67 | 1100 | 7.0379 | 0.0964 |
75
- | 6.9992 | 5.92 | 1150 | 7.0142 | 0.0975 |
76
- | 6.9382 | 6.18 | 1200 | 6.9979 | 0.0986 |
77
- | 6.956 | 6.44 | 1250 | 6.9828 | 0.0987 |
78
- | 6.9425 | 6.7 | 1300 | 6.9619 | 0.1008 |
79
- | 6.8872 | 6.96 | 1350 | 6.9468 | 0.1014 |
80
- | 6.8848 | 7.22 | 1400 | 6.9320 | 0.1024 |
81
- | 6.8578 | 7.47 | 1450 | 6.9190 | 0.1039 |
82
- | 6.8699 | 7.73 | 1500 | 6.9022 | 0.1050 |
83
- | 6.8402 | 7.99 | 1550 | 6.8910 | 0.1057 |
84
- | 6.8172 | 8.25 | 1600 | 6.8730 | 0.1069 |
85
- | 6.823 | 8.5 | 1650 | 6.8662 | 0.1073 |
86
- | 6.8028 | 8.76 | 1700 | 6.8487 | 0.1082 |
87
- | 7.3146 | 9.02 | 1750 | 6.8400 | 0.1083 |
88
- | 6.8014 | 9.28 | 1800 | 6.8303 | 0.1092 |
89
- | 6.8028 | 9.53 | 1850 | 6.8226 | 0.1088 |
90
- | 6.7817 | 9.79 | 1900 | 6.8079 | 0.1107 |
91
- | 7.28 | 10.05 | 1950 | 6.8021 | 0.1115 |
92
- | 6.7624 | 10.31 | 2000 | 6.7930 | 0.1118 |
93
- | 6.7416 | 10.56 | 2050 | 6.7868 | 0.1124 |
94
- | 6.7288 | 10.82 | 2100 | 6.7805 | 0.1133 |
95
- | 6.7468 | 11.08 | 2150 | 6.7720 | 0.1123 |
96
- | 6.7387 | 11.34 | 2200 | 6.7636 | 0.1135 |
97
- | 6.7242 | 11.6 | 2250 | 6.7557 | 0.1134 |
98
- | 6.702 | 11.85 | 2300 | 6.7496 | 0.1141 |
99
- | 6.6662 | 12.11 | 2350 | 6.7433 | 0.1150 |
100
- | 6.6781 | 12.37 | 2400 | 6.7362 | 0.1148 |
101
- | 6.6743 | 12.63 | 2450 | 6.7275 | 0.1161 |
102
- | 6.6843 | 12.88 | 2500 | 6.7247 | 0.1165 |
103
- | 6.6726 | 13.14 | 2550 | 6.7127 | 0.1173 |
104
- | 6.6656 | 13.4 | 2600 | 6.7098 | 0.1170 |
105
- | 6.6428 | 13.66 | 2650 | 6.7019 | 0.1185 |
106
- | 6.6355 | 13.91 | 2700 | 6.6979 | 0.1175 |
107
- | 6.6521 | 14.17 | 2750 | 6.6923 | 0.1188 |
108
- | 6.6735 | 14.43 | 2800 | 6.6842 | 0.1186 |
109
- | 6.6151 | 14.69 | 2850 | 6.6791 | 0.1195 |
110
- | 6.6248 | 14.94 | 2900 | 6.6752 | 0.1198 |
111
- | 6.6427 | 15.21 | 2950 | 6.6665 | 0.1207 |
112
- | 6.5947 | 15.46 | 3000 | 6.6639 | 0.1207 |
113
- | 6.6199 | 15.72 | 3050 | 6.6598 | 0.1217 |
114
- | 6.6127 | 15.98 | 3100 | 6.6593 | 0.1219 |
115
- | 6.6031 | 16.24 | 3150 | 6.6512 | 0.1226 |
116
- | 6.5742 | 16.49 | 3200 | 6.6485 | 0.1227 |
117
- | 6.621 | 16.75 | 3250 | 6.6472 | 0.1221 |
118
- | 7.0655 | 17.01 | 3300 | 6.6369 | 0.1232 |
119
- | 6.5866 | 17.27 | 3350 | 6.6376 | 0.1234 |
120
- | 6.6098 | 17.52 | 3400 | 6.6313 | 0.1252 |
121
- | 6.5676 | 17.78 | 3450 | 6.6254 | 0.1248 |
122
- | 7.0636 | 18.04 | 3500 | 6.6226 | 0.1256 |
123
- | 6.5444 | 18.3 | 3550 | 6.6164 | 0.1253 |
124
- | 6.561 | 18.55 | 3600 | 6.6157 | 0.1254 |
125
- | 6.5882 | 18.81 | 3650 | 6.6072 | 0.1257 |
126
- | 6.5518 | 19.07 | 3700 | 6.6064 | 0.1267 |
127
- | 6.5599 | 19.33 | 3750 | 6.6055 | 0.1271 |
128
- | 6.5407 | 19.59 | 3800 | 6.5987 | 0.1274 |
129
- | 6.5373 | 19.84 | 3850 | 6.5954 | 0.1280 |
130
- | 6.5381 | 20.1 | 3900 | 6.5899 | 0.1282 |
131
- | 6.5517 | 20.36 | 3950 | 6.5888 | 0.1283 |
132
- | 6.5371 | 20.62 | 4000 | 6.5854 | 0.1295 |
133
- | 6.5819 | 20.87 | 4050 | 6.5825 | 0.1282 |
134
- | 6.5425 | 21.13 | 4100 | 6.5794 | 0.1289 |
135
- | 6.5372 | 21.39 | 4150 | 6.5760 | 0.1300 |
136
- | 6.544 | 21.65 | 4200 | 6.5718 | 0.1303 |
137
- | 6.5129 | 21.9 | 4250 | 6.5660 | 0.1310 |
138
- | 6.4798 | 22.16 | 4300 | 6.5682 | 0.1305 |
139
- | 6.5556 | 22.42 | 4350 | 6.5619 | 0.1315 |
140
- | 6.4946 | 22.68 | 4400 | 6.5589 | 0.1314 |
141
- | 6.5212 | 22.93 | 4450 | 6.5593 | 0.1318 |
142
- | 6.5055 | 23.2 | 4500 | 6.5552 | 0.1311 |
143
- | 6.4693 | 23.45 | 4550 | 6.5481 | 0.1325 |
144
- | 6.4706 | 23.71 | 4600 | 6.5469 | 0.1317 |
145
- | 6.495 | 23.97 | 4650 | 6.5462 | 0.1324 |
146
- | 6.4901 | 24.23 | 4700 | 6.5414 | 0.1328 |
147
- | 6.4936 | 24.48 | 4750 | 6.5385 | 0.1334 |
148
- | 6.481 | 24.74 | 4800 | 6.5362 | 0.1331 |
149
- | 6.5186 | 25.0 | 4850 | 6.5357 | 0.1335 |
150
- | 6.4711 | 25.26 | 4900 | 6.5309 | 0.1339 |
151
- | 6.4513 | 25.51 | 4950 | 6.5284 | 0.1337 |
152
- | 6.4652 | 25.77 | 5000 | 6.5242 | 0.1343 |
153
- | 6.9335 | 26.03 | 5050 | 6.5217 | 0.1345 |
154
- | 6.4747 | 26.29 | 5100 | 6.5206 | 0.1345 |
155
- | 6.4702 | 26.54 | 5150 | 6.5201 | 0.1350 |
156
- | 6.4524 | 26.8 | 5200 | 6.5156 | 0.1352 |
157
- | 6.4225 | 27.06 | 5250 | 6.5150 | 0.1349 |
158
- | 6.4599 | 27.32 | 5300 | 6.5116 | 0.1355 |
159
- | 6.4591 | 27.58 | 5350 | 6.5098 | 0.1358 |
160
- | 6.4184 | 27.83 | 5400 | 6.5096 | 0.1353 |
161
- | 6.43 | 28.09 | 5450 | 6.5074 | 0.1361 |
162
- | 6.4604 | 28.35 | 5500 | 6.4999 | 0.1367 |
163
- | 6.4593 | 28.61 | 5550 | 6.4994 | 0.1359 |
164
- | 6.4648 | 28.86 | 5600 | 6.4981 | 0.1356 |
165
- | 6.4453 | 29.12 | 5650 | 6.4949 | 0.1374 |
166
- | 6.4275 | 29.38 | 5700 | 6.4954 | 0.1362 |
167
- | 6.4165 | 29.64 | 5750 | 6.4938 | 0.1369 |
168
- | 6.4211 | 29.89 | 5800 | 6.4911 | 0.1376 |
169
- | 6.4188 | 30.15 | 5850 | 6.4860 | 0.1374 |
170
- | 6.4337 | 30.41 | 5900 | 6.4807 | 0.1380 |
171
- | 6.4228 | 30.67 | 5950 | 6.4876 | 0.1375 |
172
- | 6.3841 | 30.92 | 6000 | 6.4811 | 0.1376 |
173
- | 6.4383 | 31.18 | 6050 | 6.4832 | 0.1379 |
174
 
175
 
176
  ### Framework versions
 
15
 
16
  This model was trained from scratch on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 2.7398
19
+ - Accuracy: 0.5144
20
 
21
  ## Model description
22
 
 
35
  ### Training hyperparameters
36
 
37
  The following hyperparameters were used during training:
38
+ - learning_rate: 0.0004
39
+ - train_batch_size: 16
40
+ - eval_batch_size: 32
41
  - seed: 42
42
+ - gradient_accumulation_steps: 32
43
+ - total_train_batch_size: 512
44
  - optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-08
45
+ - lr_scheduler_type: cosine
46
  - lr_scheduler_warmup_steps: 50
47
+ - num_epochs: 100.0
48
+ - mixed_precision_training: Native AMP
49
 
50
  ### Training results
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
53
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
54
+ | 7.8031 | 1.04 | 50 | 7.3560 | 0.0606 |
55
+ | 7.1948 | 2.08 | 100 | 6.7374 | 0.1182 |
56
+ | 6.8927 | 3.12 | 150 | 6.5022 | 0.1415 |
57
+ | 6.7339 | 4.16 | 200 | 6.4005 | 0.1483 |
58
+ | 6.6609 | 5.21 | 250 | 6.3535 | 0.1510 |
59
+ | 6.1972 | 6.25 | 300 | 6.3324 | 0.1519 |
60
+ | 6.1685 | 7.29 | 350 | 6.3029 | 0.1528 |
61
+ | 6.1302 | 8.33 | 400 | 6.2828 | 0.1521 |
62
+ | 6.093 | 9.37 | 450 | 6.2568 | 0.1536 |
63
+ | 6.0543 | 10.41 | 500 | 6.2430 | 0.1544 |
64
+ | 6.0479 | 11.45 | 550 | 6.2346 | 0.1541 |
65
+ | 6.0372 | 12.49 | 600 | 6.2232 | 0.1546 |
66
+ | 6.0127 | 13.53 | 650 | 6.2139 | 0.1541 |
67
+ | 5.968 | 14.58 | 700 | 6.2053 | 0.1547 |
68
+ | 5.9635 | 15.62 | 750 | 6.1996 | 0.1549 |
69
+ | 5.9479 | 16.66 | 800 | 6.1953 | 0.1548 |
70
+ | 5.9371 | 17.7 | 850 | 6.1887 | 0.1545 |
71
+ | 5.9046 | 18.74 | 900 | 6.1613 | 0.1545 |
72
+ | 5.8368 | 19.78 | 950 | 6.0952 | 0.1557 |
73
+ | 5.7914 | 20.82 | 1000 | 6.0330 | 0.1569 |
74
+ | 5.7026 | 21.86 | 1050 | 5.9430 | 0.1612 |
75
+ | 5.491 | 22.9 | 1100 | 5.6100 | 0.1974 |
76
+ | 4.9289 | 23.95 | 1150 | 4.9607 | 0.2702 |
77
+ | 4.5214 | 24.99 | 1200 | 4.5795 | 0.3051 |
78
+ | 4.5663 | 26.04 | 1250 | 4.3454 | 0.3265 |
79
+ | 4.3717 | 27.08 | 1300 | 4.1738 | 0.3412 |
80
+ | 4.1483 | 28.12 | 1350 | 4.0336 | 0.3555 |
81
+ | 3.9988 | 29.16 | 1400 | 3.9180 | 0.3677 |
82
+ | 3.8695 | 30.21 | 1450 | 3.8108 | 0.3782 |
83
+ | 3.5017 | 31.25 | 1500 | 3.7240 | 0.3879 |
84
+ | 3.4311 | 32.29 | 1550 | 3.6426 | 0.3974 |
85
+ | 3.3517 | 33.33 | 1600 | 3.5615 | 0.4068 |
86
+ | 3.2856 | 34.37 | 1650 | 3.4915 | 0.4156 |
87
+ | 3.227 | 35.41 | 1700 | 3.4179 | 0.4255 |
88
+ | 3.1675 | 36.45 | 1750 | 3.3636 | 0.4325 |
89
+ | 3.0908 | 37.49 | 1800 | 3.3083 | 0.4394 |
90
+ | 3.0561 | 38.53 | 1850 | 3.2572 | 0.4473 |
91
+ | 3.0139 | 39.58 | 1900 | 3.2159 | 0.4525 |
92
+ | 2.9837 | 40.62 | 1950 | 3.1789 | 0.4575 |
93
+ | 2.9387 | 41.66 | 2000 | 3.1431 | 0.4618 |
94
+ | 2.9034 | 42.7 | 2050 | 3.1163 | 0.4654 |
95
+ | 2.8822 | 43.74 | 2100 | 3.0842 | 0.4694 |
96
+ | 2.836 | 44.78 | 2150 | 3.0583 | 0.4727 |
97
+ | 2.8129 | 45.82 | 2200 | 3.0359 | 0.4760 |
98
+ | 2.7733 | 46.86 | 2250 | 3.0173 | 0.4776 |
99
+ | 2.7589 | 47.9 | 2300 | 2.9978 | 0.4812 |
100
+ | 2.7378 | 48.95 | 2350 | 2.9788 | 0.4831 |
101
+ | 2.7138 | 49.99 | 2400 | 2.9674 | 0.4844 |
102
+ | 2.8692 | 51.04 | 2450 | 2.9476 | 0.4874 |
103
+ | 2.8462 | 52.08 | 2500 | 2.9342 | 0.4893 |
104
+ | 2.8312 | 53.12 | 2550 | 2.9269 | 0.4900 |
105
+ | 2.7834 | 54.16 | 2600 | 2.9111 | 0.4917 |
106
+ | 2.7822 | 55.21 | 2650 | 2.8987 | 0.4934 |
107
+ | 2.584 | 56.25 | 2700 | 2.8844 | 0.4949 |
108
+ | 2.5668 | 57.29 | 2750 | 2.8808 | 0.4965 |
109
+ | 2.5536 | 58.33 | 2800 | 2.8640 | 0.4982 |
110
+ | 2.5403 | 59.37 | 2850 | 2.8606 | 0.4982 |
111
+ | 2.5294 | 60.41 | 2900 | 2.8441 | 0.5008 |
112
+ | 2.513 | 61.45 | 2950 | 2.8402 | 0.5013 |
113
+ | 2.5105 | 62.49 | 3000 | 2.8316 | 0.5022 |
114
+ | 2.4897 | 63.53 | 3050 | 2.8237 | 0.5027 |
115
+ | 2.4974 | 64.58 | 3100 | 2.8187 | 0.5040 |
116
+ | 2.4799 | 65.62 | 3150 | 2.8129 | 0.5044 |
117
+ | 2.4741 | 66.66 | 3200 | 2.8056 | 0.5057 |
118
+ | 2.4582 | 67.7 | 3250 | 2.8025 | 0.5061 |
119
+ | 2.4389 | 68.74 | 3300 | 2.7913 | 0.5076 |
120
+ | 2.4539 | 69.78 | 3350 | 2.7881 | 0.5072 |
121
+ | 2.4252 | 70.82 | 3400 | 2.7884 | 0.5082 |
122
+ | 2.4287 | 71.86 | 3450 | 2.7784 | 0.5093 |
123
+ | 2.4131 | 72.9 | 3500 | 2.7782 | 0.5099 |
124
+ | 2.4016 | 73.95 | 3550 | 2.7724 | 0.5098 |
125
+ | 2.3998 | 74.99 | 3600 | 2.7659 | 0.5111 |
126
+ | 2.5475 | 76.04 | 3650 | 2.7650 | 0.5108 |
127
+ | 2.5443 | 77.08 | 3700 | 2.7620 | 0.5117 |
128
+ | 2.5381 | 78.12 | 3750 | 2.7631 | 0.5115 |
129
+ | 2.5269 | 79.16 | 3800 | 2.7578 | 0.5122 |
130
+ | 2.5288 | 80.21 | 3850 | 2.7540 | 0.5124 |
131
+ | 2.3669 | 81.25 | 3900 | 2.7529 | 0.5125 |
132
+ | 2.3631 | 82.29 | 3950 | 2.7498 | 0.5132 |
133
+ | 2.3499 | 83.33 | 4000 | 2.7454 | 0.5136 |
134
+ | 2.3726 | 84.37 | 4050 | 2.7446 | 0.5141 |
135
+ | 2.3411 | 85.41 | 4100 | 2.7403 | 0.5144 |
136
+ | 2.3321 | 86.45 | 4150 | 2.7372 | 0.5146 |
137
+ | 2.3456 | 87.49 | 4200 | 2.7389 | 0.5146 |
138
+ | 2.3372 | 88.53 | 4250 | 2.7384 | 0.5151 |
139
+ | 2.343 | 89.58 | 4300 | 2.7398 | 0.5144 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
140
 
141
 
142
  ### Framework versions