<!-- 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. -->
bert-base-uncased_winobias_classifieronly
This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6933
- Accuracy: 0.5114
- Tp: 0.3112
- Tn: 0.2001
- Fp: 0.2999
- Fn: 0.1888
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Tp | Tn | Fp | Fn |
---|---|---|---|---|---|---|---|---|
0.7029 | 0.8 | 20 | 0.6940 | 0.4867 | 0.1951 | 0.2917 | 0.2083 | 0.3049 |
0.6922 | 1.6 | 40 | 0.6950 | 0.5082 | 0.4293 | 0.0789 | 0.4211 | 0.0707 |
0.6953 | 2.4 | 60 | 0.6955 | 0.5019 | 0.4482 | 0.0537 | 0.4463 | 0.0518 |
0.6955 | 3.2 | 80 | 0.6940 | 0.5038 | 0.3365 | 0.1673 | 0.3327 | 0.1635 |
0.7066 | 4.0 | 100 | 0.6941 | 0.5 | 0.3630 | 0.1370 | 0.3630 | 0.1370 |
0.6914 | 4.8 | 120 | 0.6939 | 0.4880 | 0.1824 | 0.3056 | 0.1944 | 0.3176 |
0.69 | 5.6 | 140 | 0.6943 | 0.5025 | 0.3977 | 0.1048 | 0.3952 | 0.1023 |
0.6975 | 6.4 | 160 | 0.6938 | 0.5032 | 0.2342 | 0.2689 | 0.2311 | 0.2658 |
0.6935 | 7.2 | 180 | 0.6938 | 0.4968 | 0.2247 | 0.2721 | 0.2279 | 0.2753 |
0.7033 | 8.0 | 200 | 0.6941 | 0.5019 | 0.3693 | 0.1326 | 0.3674 | 0.1307 |
0.7004 | 8.8 | 220 | 0.6938 | 0.5051 | 0.3138 | 0.1913 | 0.3087 | 0.1862 |
0.6978 | 9.6 | 240 | 0.6939 | 0.5019 | 0.2588 | 0.2431 | 0.2569 | 0.2412 |
0.7005 | 10.4 | 260 | 0.6940 | 0.5063 | 0.3289 | 0.1774 | 0.3226 | 0.1711 |
0.7006 | 11.2 | 280 | 0.6940 | 0.5 | 0.2355 | 0.2645 | 0.2355 | 0.2645 |
0.6978 | 12.0 | 300 | 0.6942 | 0.5051 | 0.3649 | 0.1402 | 0.3598 | 0.1351 |
0.6938 | 12.8 | 320 | 0.6940 | 0.5025 | 0.2582 | 0.2443 | 0.2557 | 0.2418 |
0.7024 | 13.6 | 340 | 0.6946 | 0.4962 | 0.1199 | 0.3763 | 0.1237 | 0.3801 |
0.7002 | 14.4 | 360 | 0.6944 | 0.4975 | 0.4110 | 0.0865 | 0.4135 | 0.0890 |
0.6952 | 15.2 | 380 | 0.6939 | 0.5076 | 0.2753 | 0.2323 | 0.2677 | 0.2247 |
0.6936 | 16.0 | 400 | 0.6943 | 0.4848 | 0.1496 | 0.3352 | 0.1648 | 0.3504 |
0.6983 | 16.8 | 420 | 0.6940 | 0.4962 | 0.1970 | 0.2992 | 0.2008 | 0.3030 |
0.6944 | 17.6 | 440 | 0.6939 | 0.4956 | 0.2241 | 0.2715 | 0.2285 | 0.2759 |
0.6984 | 18.4 | 460 | 0.6938 | 0.5120 | 0.3074 | 0.2045 | 0.2955 | 0.1926 |
0.6933 | 19.2 | 480 | 0.6937 | 0.5013 | 0.3504 | 0.1509 | 0.3491 | 0.1496 |
0.6918 | 20.0 | 500 | 0.6937 | 0.5032 | 0.3466 | 0.1566 | 0.3434 | 0.1534 |
0.6989 | 20.8 | 520 | 0.6936 | 0.4987 | 0.3441 | 0.1547 | 0.3453 | 0.1559 |
0.6939 | 21.6 | 540 | 0.6936 | 0.5013 | 0.3327 | 0.1686 | 0.3314 | 0.1673 |
0.6982 | 22.4 | 560 | 0.6936 | 0.5095 | 0.3125 | 0.1970 | 0.3030 | 0.1875 |
0.6924 | 23.2 | 580 | 0.6937 | 0.5044 | 0.3674 | 0.1370 | 0.3630 | 0.1326 |
0.6975 | 24.0 | 600 | 0.6936 | 0.5013 | 0.3346 | 0.1667 | 0.3333 | 0.1654 |
0.6934 | 24.8 | 620 | 0.6937 | 0.5013 | 0.3718 | 0.1294 | 0.3706 | 0.1282 |
0.6929 | 25.6 | 640 | 0.6934 | 0.4975 | 0.2064 | 0.2910 | 0.2090 | 0.2936 |
0.7 | 26.4 | 660 | 0.6934 | 0.5114 | 0.2645 | 0.2468 | 0.2532 | 0.2355 |
0.6994 | 27.2 | 680 | 0.6933 | 0.5069 | 0.2879 | 0.2191 | 0.2809 | 0.2121 |
0.6992 | 28.0 | 700 | 0.6933 | 0.5145 | 0.3157 | 0.1989 | 0.3011 | 0.1843 |
0.6965 | 28.8 | 720 | 0.6933 | 0.5095 | 0.2828 | 0.2266 | 0.2734 | 0.2172 |
0.6978 | 29.6 | 740 | 0.6936 | 0.5019 | 0.3807 | 0.1212 | 0.3788 | 0.1193 |
0.7004 | 30.4 | 760 | 0.6934 | 0.5019 | 0.3251 | 0.1768 | 0.3232 | 0.1749 |
0.695 | 31.2 | 780 | 0.6934 | 0.5120 | 0.2866 | 0.2254 | 0.2746 | 0.2134 |
0.6949 | 32.0 | 800 | 0.6935 | 0.5025 | 0.3491 | 0.1534 | 0.3466 | 0.1509 |
0.6929 | 32.8 | 820 | 0.6934 | 0.5019 | 0.3472 | 0.1547 | 0.3453 | 0.1528 |
0.6953 | 33.6 | 840 | 0.6933 | 0.5126 | 0.2746 | 0.2380 | 0.2620 | 0.2254 |
0.6967 | 34.4 | 860 | 0.6935 | 0.5038 | 0.3687 | 0.1351 | 0.3649 | 0.1313 |
0.6921 | 35.2 | 880 | 0.6937 | 0.5051 | 0.4040 | 0.1010 | 0.3990 | 0.0960 |
0.6973 | 36.0 | 900 | 0.6936 | 0.5013 | 0.3788 | 0.1225 | 0.3775 | 0.1212 |
0.7016 | 36.8 | 920 | 0.6933 | 0.5038 | 0.3283 | 0.1755 | 0.3245 | 0.1717 |
0.6927 | 37.6 | 940 | 0.6934 | 0.5013 | 0.3624 | 0.1389 | 0.3611 | 0.1376 |
0.6966 | 38.4 | 960 | 0.6934 | 0.5038 | 0.3422 | 0.1616 | 0.3384 | 0.1578 |
0.6981 | 39.2 | 980 | 0.6933 | 0.5038 | 0.3207 | 0.1831 | 0.3169 | 0.1793 |
0.6964 | 40.0 | 1000 | 0.6934 | 0.5025 | 0.3403 | 0.1622 | 0.3378 | 0.1597 |
0.6965 | 40.8 | 1020 | 0.6933 | 0.5025 | 0.3207 | 0.1818 | 0.3182 | 0.1793 |
0.6981 | 41.6 | 1040 | 0.6933 | 0.5177 | 0.3087 | 0.2090 | 0.2910 | 0.1913 |
0.7021 | 42.4 | 1060 | 0.6933 | 0.5164 | 0.2828 | 0.2336 | 0.2664 | 0.2172 |
0.6948 | 43.2 | 1080 | 0.6933 | 0.5063 | 0.2544 | 0.2519 | 0.2481 | 0.2456 |
0.695 | 44.0 | 1100 | 0.6933 | 0.5044 | 0.2519 | 0.2525 | 0.2475 | 0.2481 |
0.692 | 44.8 | 1120 | 0.6933 | 0.5158 | 0.2923 | 0.2235 | 0.2765 | 0.2077 |
0.6925 | 45.6 | 1140 | 0.6933 | 0.5158 | 0.3062 | 0.2096 | 0.2904 | 0.1938 |
0.6983 | 46.4 | 1160 | 0.6933 | 0.5063 | 0.3213 | 0.1850 | 0.3150 | 0.1787 |
0.6953 | 47.2 | 1180 | 0.6933 | 0.5044 | 0.3258 | 0.1787 | 0.3213 | 0.1742 |
0.693 | 48.0 | 1200 | 0.6933 | 0.5063 | 0.3201 | 0.1862 | 0.3138 | 0.1799 |
0.6944 | 48.8 | 1220 | 0.6933 | 0.5095 | 0.3188 | 0.1907 | 0.3093 | 0.1812 |
0.6939 | 49.6 | 1240 | 0.6933 | 0.5114 | 0.3112 | 0.2001 | 0.2999 | 0.1888 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1
- Datasets 2.10.1
- Tokenizers 0.13.2