<!-- 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. -->
multiberts-seed_2-step_2000k_winobias_classifieronly
This model is a fine-tuned version of google/multiberts-seed_2-step_2000k on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6937
- Accuracy: 0.4943
- Tp: 0.1629
- Tn: 0.3314
- Fp: 0.1686
- Fn: 0.3371
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 |
---|---|---|---|---|---|---|---|---|
0.7201 | 0.8 | 20 | 0.7046 | 0.5044 | 0.0884 | 0.4160 | 0.0840 | 0.4116 |
0.7141 | 1.6 | 40 | 0.6997 | 0.5025 | 0.2001 | 0.3024 | 0.1976 | 0.2999 |
0.7113 | 2.4 | 60 | 0.6979 | 0.5063 | 0.2936 | 0.2128 | 0.2872 | 0.2064 |
0.7073 | 3.2 | 80 | 0.6967 | 0.5038 | 0.1225 | 0.3813 | 0.1187 | 0.3775 |
0.6922 | 4.0 | 100 | 0.6950 | 0.5019 | 0.1616 | 0.3403 | 0.1597 | 0.3384 |
0.7025 | 4.8 | 120 | 0.6953 | 0.5076 | 0.1313 | 0.3763 | 0.1237 | 0.3687 |
0.7029 | 5.6 | 140 | 0.6947 | 0.5019 | 0.2986 | 0.2033 | 0.2967 | 0.2014 |
0.6974 | 6.4 | 160 | 0.6952 | 0.5038 | 0.1092 | 0.3946 | 0.1054 | 0.3908 |
0.6992 | 7.2 | 180 | 0.6948 | 0.5088 | 0.1275 | 0.3813 | 0.1187 | 0.3725 |
0.6944 | 8.0 | 200 | 0.6939 | 0.4956 | 0.2557 | 0.2399 | 0.2601 | 0.2443 |
0.6953 | 8.8 | 220 | 0.6940 | 0.4912 | 0.1824 | 0.3087 | 0.1913 | 0.3176 |
0.6994 | 9.6 | 240 | 0.6942 | 0.4949 | 0.1503 | 0.3447 | 0.1553 | 0.3497 |
0.6955 | 10.4 | 260 | 0.6939 | 0.4949 | 0.2405 | 0.2544 | 0.2456 | 0.2595 |
0.6993 | 11.2 | 280 | 0.6942 | 0.5006 | 0.1446 | 0.3561 | 0.1439 | 0.3554 |
0.6925 | 12.0 | 300 | 0.6940 | 0.4975 | 0.1616 | 0.3359 | 0.1641 | 0.3384 |
0.6985 | 12.8 | 320 | 0.6938 | 0.4905 | 0.2008 | 0.2898 | 0.2102 | 0.2992 |
0.7014 | 13.6 | 340 | 0.6951 | 0.5051 | 0.0821 | 0.4230 | 0.0770 | 0.4179 |
0.6947 | 14.4 | 360 | 0.6939 | 0.4912 | 0.3150 | 0.1761 | 0.3239 | 0.1850 |
0.698 | 15.2 | 380 | 0.6940 | 0.5006 | 0.1654 | 0.3352 | 0.1648 | 0.3346 |
0.6912 | 16.0 | 400 | 0.6946 | 0.5032 | 0.1073 | 0.3958 | 0.1042 | 0.3927 |
0.6929 | 16.8 | 420 | 0.6946 | 0.5 | 0.1035 | 0.3965 | 0.1035 | 0.3965 |
0.6946 | 17.6 | 440 | 0.6938 | 0.4994 | 0.1951 | 0.3043 | 0.1957 | 0.3049 |
0.6955 | 18.4 | 460 | 0.6937 | 0.4962 | 0.2481 | 0.2481 | 0.2519 | 0.2519 |
0.7 | 19.2 | 480 | 0.6938 | 0.4994 | 0.1894 | 0.3100 | 0.1900 | 0.3106 |
0.6947 | 20.0 | 500 | 0.6938 | 0.4994 | 0.2008 | 0.2986 | 0.2014 | 0.2992 |
0.6978 | 20.8 | 520 | 0.6937 | 0.4937 | 0.2462 | 0.2475 | 0.2525 | 0.2538 |
0.7004 | 21.6 | 540 | 0.6937 | 0.4962 | 0.2677 | 0.2285 | 0.2715 | 0.2323 |
0.6977 | 22.4 | 560 | 0.6937 | 0.4975 | 0.2620 | 0.2355 | 0.2645 | 0.2380 |
0.6933 | 23.2 | 580 | 0.6937 | 0.4968 | 0.2342 | 0.2626 | 0.2374 | 0.2658 |
0.6991 | 24.0 | 600 | 0.6938 | 0.5 | 0.1824 | 0.3176 | 0.1824 | 0.3176 |
0.6961 | 24.8 | 620 | 0.6937 | 0.4987 | 0.2140 | 0.2847 | 0.2153 | 0.2860 |
0.7054 | 25.6 | 640 | 0.6944 | 0.5032 | 0.1029 | 0.4003 | 0.0997 | 0.3971 |
0.6991 | 26.4 | 660 | 0.6942 | 0.4943 | 0.1181 | 0.3763 | 0.1237 | 0.3819 |
0.702 | 27.2 | 680 | 0.6943 | 0.5013 | 0.1004 | 0.4009 | 0.0991 | 0.3996 |
0.6968 | 28.0 | 700 | 0.6941 | 0.4918 | 0.1206 | 0.3712 | 0.1288 | 0.3794 |
0.6939 | 28.8 | 720 | 0.6941 | 0.4912 | 0.1136 | 0.3775 | 0.1225 | 0.3864 |
0.6956 | 29.6 | 740 | 0.6936 | 0.5019 | 0.2361 | 0.2658 | 0.2342 | 0.2639 |
0.6956 | 30.4 | 760 | 0.6936 | 0.4968 | 0.2172 | 0.2797 | 0.2203 | 0.2828 |
0.6987 | 31.2 | 780 | 0.6937 | 0.4924 | 0.1679 | 0.3245 | 0.1755 | 0.3321 |
0.6935 | 32.0 | 800 | 0.6936 | 0.4912 | 0.1970 | 0.2942 | 0.2058 | 0.3030 |
0.6967 | 32.8 | 820 | 0.6936 | 0.4937 | 0.1913 | 0.3024 | 0.1976 | 0.3087 |
0.6982 | 33.6 | 840 | 0.6941 | 0.4968 | 0.1288 | 0.3681 | 0.1319 | 0.3712 |
0.7007 | 34.4 | 860 | 0.6937 | 0.4949 | 0.1824 | 0.3125 | 0.1875 | 0.3176 |
0.7 | 35.2 | 880 | 0.6936 | 0.4949 | 0.2197 | 0.2753 | 0.2247 | 0.2803 |
0.6904 | 36.0 | 900 | 0.6935 | 0.5025 | 0.2431 | 0.2595 | 0.2405 | 0.2569 |
0.6945 | 36.8 | 920 | 0.6936 | 0.4937 | 0.2096 | 0.2841 | 0.2159 | 0.2904 |
0.7025 | 37.6 | 940 | 0.6935 | 0.5019 | 0.2519 | 0.25 | 0.25 | 0.2481 |
0.6969 | 38.4 | 960 | 0.6935 | 0.4994 | 0.2235 | 0.2759 | 0.2241 | 0.2765 |
0.7038 | 39.2 | 980 | 0.6936 | 0.4949 | 0.1818 | 0.3131 | 0.1869 | 0.3182 |
0.698 | 40.0 | 1000 | 0.6937 | 0.4931 | 0.1736 | 0.3194 | 0.1806 | 0.3264 |
0.6973 | 40.8 | 1020 | 0.6938 | 0.5013 | 0.1540 | 0.3472 | 0.1528 | 0.3460 |
0.6964 | 41.6 | 1040 | 0.6939 | 0.5032 | 0.1408 | 0.3624 | 0.1376 | 0.3592 |
0.6999 | 42.4 | 1060 | 0.6939 | 0.5 | 0.1370 | 0.3630 | 0.1370 | 0.3630 |
0.7002 | 43.2 | 1080 | 0.6939 | 0.5006 | 0.1376 | 0.3630 | 0.1370 | 0.3624 |
0.6939 | 44.0 | 1100 | 0.6940 | 0.4956 | 0.1225 | 0.3731 | 0.1269 | 0.3775 |
0.6984 | 44.8 | 1120 | 0.6939 | 0.4994 | 0.1338 | 0.3655 | 0.1345 | 0.3662 |
0.6946 | 45.6 | 1140 | 0.6939 | 0.5019 | 0.1395 | 0.3624 | 0.1376 | 0.3605 |
0.6972 | 46.4 | 1160 | 0.6937 | 0.4962 | 0.1616 | 0.3346 | 0.1654 | 0.3384 |
0.694 | 47.2 | 1180 | 0.6937 | 0.4905 | 0.1679 | 0.3226 | 0.1774 | 0.3321 |
0.6974 | 48.0 | 1200 | 0.6937 | 0.4886 | 0.1648 | 0.3239 | 0.1761 | 0.3352 |
0.6956 | 48.8 | 1220 | 0.6937 | 0.4893 | 0.1648 | 0.3245 | 0.1755 | 0.3352 |
0.7032 | 49.6 | 1240 | 0.6937 | 0.4943 | 0.1629 | 0.3314 | 0.1686 | 0.3371 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1
- Datasets 2.10.1
- Tokenizers 0.13.2