<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
IMDB_DistilBERT_5EE
This model is a fine-tuned version of distilbert-base-uncased on the imdb dataset. It achieves the following results on the evaluation set:
- Loss: 0.2023
- Accuracy: 0.94
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.6748 | 0.03 | 50 | 0.5955 | 0.88 |
0.4404 | 0.06 | 100 | 0.2853 | 0.9 |
0.3065 | 0.1 | 150 | 0.2208 | 0.9 |
0.3083 | 0.13 | 200 | 0.2023 | 0.9333 |
0.2922 | 0.16 | 250 | 0.1530 | 0.94 |
0.2761 | 0.19 | 300 | 0.2035 | 0.9267 |
0.2145 | 0.22 | 350 | 0.2450 | 0.9 |
0.258 | 0.26 | 400 | 0.1680 | 0.9267 |
0.2702 | 0.29 | 450 | 0.1607 | 0.9333 |
0.2587 | 0.32 | 500 | 0.1496 | 0.9467 |
0.2822 | 0.35 | 550 | 0.1405 | 0.9333 |
0.2538 | 0.38 | 600 | 0.1396 | 0.9467 |
0.2707 | 0.42 | 650 | 0.1626 | 0.9333 |
0.2408 | 0.45 | 700 | 0.1623 | 0.9067 |
0.2531 | 0.48 | 750 | 0.1300 | 0.9467 |
0.2014 | 0.51 | 800 | 0.1529 | 0.9333 |
0.2454 | 0.54 | 850 | 0.1365 | 0.94 |
0.2282 | 0.58 | 900 | 0.1447 | 0.9533 |
0.2554 | 0.61 | 950 | 0.1321 | 0.9467 |
0.24 | 0.64 | 1000 | 0.1256 | 0.9467 |
0.2239 | 0.67 | 1050 | 0.1290 | 0.9467 |
0.2865 | 0.7 | 1100 | 0.1288 | 0.9667 |
0.2456 | 0.74 | 1150 | 0.1299 | 0.9533 |
0.2407 | 0.77 | 1200 | 0.1565 | 0.9267 |
0.2256 | 0.8 | 1250 | 0.1262 | 0.96 |
0.238 | 0.83 | 1300 | 0.1599 | 0.9333 |
0.2151 | 0.86 | 1350 | 0.1252 | 0.9333 |
0.187 | 0.9 | 1400 | 0.1132 | 0.9467 |
0.2218 | 0.93 | 1450 | 0.1030 | 0.9533 |
0.2371 | 0.96 | 1500 | 0.1036 | 0.9467 |
0.2264 | 0.99 | 1550 | 0.1041 | 0.9467 |
0.2159 | 1.02 | 1600 | 0.1338 | 0.9267 |
0.1773 | 1.06 | 1650 | 0.1218 | 0.94 |
0.1381 | 1.09 | 1700 | 0.1593 | 0.94 |
0.1582 | 1.12 | 1750 | 0.1445 | 0.9533 |
0.1921 | 1.15 | 1800 | 0.1355 | 0.94 |
0.206 | 1.18 | 1850 | 0.1511 | 0.9467 |
0.1679 | 1.22 | 1900 | 0.1394 | 0.94 |
0.1691 | 1.25 | 1950 | 0.1403 | 0.9333 |
0.2301 | 1.28 | 2000 | 0.1169 | 0.9467 |
0.1764 | 1.31 | 2050 | 0.1507 | 0.9333 |
0.1772 | 1.34 | 2100 | 0.1148 | 0.96 |
0.1749 | 1.38 | 2150 | 0.1203 | 0.94 |
0.1912 | 1.41 | 2200 | 0.1037 | 0.94 |
0.1614 | 1.44 | 2250 | 0.1006 | 0.9533 |
0.1975 | 1.47 | 2300 | 0.0985 | 0.9533 |
0.1843 | 1.5 | 2350 | 0.0922 | 0.9533 |
0.1764 | 1.54 | 2400 | 0.1259 | 0.9467 |
0.1855 | 1.57 | 2450 | 0.1243 | 0.96 |
0.1272 | 1.6 | 2500 | 0.2107 | 0.9267 |
0.241 | 1.63 | 2550 | 0.1142 | 0.9533 |
0.1584 | 1.66 | 2600 | 0.1194 | 0.9467 |
0.1568 | 1.7 | 2650 | 0.1196 | 0.9533 |
0.1896 | 1.73 | 2700 | 0.1311 | 0.9533 |
0.143 | 1.76 | 2750 | 0.1140 | 0.9533 |
0.227 | 1.79 | 2800 | 0.1482 | 0.9333 |
0.1404 | 1.82 | 2850 | 0.1366 | 0.94 |
0.1865 | 1.86 | 2900 | 0.1174 | 0.94 |
0.1659 | 1.89 | 2950 | 0.1189 | 0.94 |
0.1882 | 1.92 | 3000 | 0.1144 | 0.9467 |
0.1403 | 1.95 | 3050 | 0.1358 | 0.94 |
0.2193 | 1.98 | 3100 | 0.1092 | 0.9533 |
0.1392 | 2.02 | 3150 | 0.1278 | 0.9267 |
0.1292 | 2.05 | 3200 | 0.1186 | 0.96 |
0.0939 | 2.08 | 3250 | 0.1183 | 0.94 |
0.1356 | 2.11 | 3300 | 0.1939 | 0.94 |
0.1175 | 2.14 | 3350 | 0.1499 | 0.94 |
0.1285 | 2.18 | 3400 | 0.1538 | 0.94 |
0.1018 | 2.21 | 3450 | 0.1796 | 0.9333 |
0.1342 | 2.24 | 3500 | 0.1540 | 0.94 |
0.17 | 2.27 | 3550 | 0.1261 | 0.94 |
0.1548 | 2.3 | 3600 | 0.1375 | 0.9267 |
0.1415 | 2.34 | 3650 | 0.1264 | 0.9333 |
0.1096 | 2.37 | 3700 | 0.1252 | 0.9333 |
0.1001 | 2.4 | 3750 | 0.1546 | 0.94 |
0.0934 | 2.43 | 3800 | 0.1534 | 0.94 |
0.1287 | 2.46 | 3850 | 0.1735 | 0.9333 |
0.0872 | 2.5 | 3900 | 0.1475 | 0.9467 |
0.0994 | 2.53 | 3950 | 0.1735 | 0.9467 |
0.1558 | 2.56 | 4000 | 0.1585 | 0.9467 |
0.1517 | 2.59 | 4050 | 0.2021 | 0.9333 |
0.1246 | 2.62 | 4100 | 0.1594 | 0.9267 |
0.1228 | 2.66 | 4150 | 0.1338 | 0.9533 |
0.1064 | 2.69 | 4200 | 0.1421 | 0.9467 |
0.1466 | 2.72 | 4250 | 0.1383 | 0.9467 |
0.1243 | 2.75 | 4300 | 0.1604 | 0.9533 |
0.1434 | 2.78 | 4350 | 0.1736 | 0.9333 |
0.1127 | 2.82 | 4400 | 0.1909 | 0.9267 |
0.0908 | 2.85 | 4450 | 0.1958 | 0.9333 |
0.1134 | 2.88 | 4500 | 0.1596 | 0.94 |
0.1345 | 2.91 | 4550 | 0.1604 | 0.9533 |
0.1913 | 2.94 | 4600 | 0.1852 | 0.9267 |
0.1382 | 2.98 | 4650 | 0.1852 | 0.9333 |
0.1109 | 3.01 | 4700 | 0.1905 | 0.9333 |
0.1144 | 3.04 | 4750 | 0.1655 | 0.94 |
0.074 | 3.07 | 4800 | 0.2034 | 0.9333 |
0.0926 | 3.1 | 4850 | 0.1929 | 0.94 |
0.0911 | 3.13 | 4900 | 0.1703 | 0.9333 |
0.0933 | 3.17 | 4950 | 0.1826 | 0.9333 |
0.1003 | 3.2 | 5000 | 0.1716 | 0.94 |
0.0889 | 3.23 | 5050 | 0.1843 | 0.9267 |
0.0841 | 3.26 | 5100 | 0.1670 | 0.94 |
0.0918 | 3.29 | 5150 | 0.1595 | 0.9467 |
0.0795 | 3.33 | 5200 | 0.1504 | 0.96 |
0.0978 | 3.36 | 5250 | 0.1317 | 0.96 |
0.1202 | 3.39 | 5300 | 0.1641 | 0.9533 |
0.0935 | 3.42 | 5350 | 0.1473 | 0.96 |
0.0673 | 3.45 | 5400 | 0.1684 | 0.9533 |
0.0729 | 3.49 | 5450 | 0.1414 | 0.9533 |
0.077 | 3.52 | 5500 | 0.1669 | 0.9533 |
0.1264 | 3.55 | 5550 | 0.1364 | 0.96 |
0.1282 | 3.58 | 5600 | 0.1575 | 0.9467 |
0.0553 | 3.61 | 5650 | 0.1440 | 0.9467 |
0.0953 | 3.65 | 5700 | 0.1526 | 0.9533 |
0.0886 | 3.68 | 5750 | 0.1633 | 0.94 |
0.0901 | 3.71 | 5800 | 0.1704 | 0.9467 |
0.0986 | 3.74 | 5850 | 0.1674 | 0.94 |
0.0849 | 3.77 | 5900 | 0.1989 | 0.9333 |
0.0815 | 3.81 | 5950 | 0.1942 | 0.94 |
0.0973 | 3.84 | 6000 | 0.1611 | 0.94 |
0.0599 | 3.87 | 6050 | 0.1807 | 0.9267 |
0.1068 | 3.9 | 6100 | 0.1966 | 0.94 |
0.0889 | 3.93 | 6150 | 0.1979 | 0.9333 |
0.0854 | 3.97 | 6200 | 0.2012 | 0.9333 |
0.1207 | 4.0 | 6250 | 0.1983 | 0.9333 |
0.0735 | 4.03 | 6300 | 0.1795 | 0.94 |
0.1148 | 4.06 | 6350 | 0.1966 | 0.94 |
0.0725 | 4.09 | 6400 | 0.2290 | 0.94 |
0.0576 | 4.13 | 6450 | 0.1936 | 0.9333 |
0.0477 | 4.16 | 6500 | 0.2090 | 0.9333 |
0.0722 | 4.19 | 6550 | 0.1878 | 0.9333 |
0.0936 | 4.22 | 6600 | 0.2087 | 0.94 |
0.0715 | 4.25 | 6650 | 0.2040 | 0.94 |
0.0586 | 4.29 | 6700 | 0.1862 | 0.9333 |
0.0548 | 4.32 | 6750 | 0.1801 | 0.9267 |
0.0527 | 4.35 | 6800 | 0.1912 | 0.9333 |
0.0813 | 4.38 | 6850 | 0.1941 | 0.9333 |
0.0531 | 4.41 | 6900 | 0.1932 | 0.9267 |
0.0606 | 4.45 | 6950 | 0.2195 | 0.94 |
0.1213 | 4.48 | 7000 | 0.1975 | 0.9333 |
0.0807 | 4.51 | 7050 | 0.1915 | 0.9333 |
0.076 | 4.54 | 7100 | 0.1987 | 0.9333 |
0.0595 | 4.57 | 7150 | 0.2052 | 0.9333 |
0.0832 | 4.61 | 7200 | 0.2039 | 0.9333 |
0.0657 | 4.64 | 7250 | 0.2186 | 0.94 |
0.0684 | 4.67 | 7300 | 0.2063 | 0.94 |
0.0429 | 4.7 | 7350 | 0.2056 | 0.94 |
0.0531 | 4.73 | 7400 | 0.2139 | 0.94 |
0.0556 | 4.77 | 7450 | 0.2153 | 0.94 |
0.0824 | 4.8 | 7500 | 0.2010 | 0.9333 |
0.039 | 4.83 | 7550 | 0.2079 | 0.94 |
0.068 | 4.86 | 7600 | 0.2140 | 0.94 |
0.065 | 4.89 | 7650 | 0.2108 | 0.94 |
0.0359 | 4.93 | 7700 | 0.2058 | 0.94 |
0.0592 | 4.96 | 7750 | 0.2029 | 0.94 |
0.0793 | 4.99 | 7800 | 0.2023 | 0.94 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.1