<!-- 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_1_winobias_classifieronly
This model is a fine-tuned version of google/multiberts-seed_1 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6936
- Accuracy: 0.5114
- Tp: 0.2734
- Tn: 0.2380
- Fp: 0.2620
- Fn: 0.2266
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 | 0.6948 | 0.5019 | 0.1951 | 0.3068 | 0.1932 | 0.3049 |
0.6937 | 1.6 | 40 | 0.6952 | 0.4931 | 0.3390 | 0.1540 | 0.3460 | 0.1610 |
0.6974 | 2.4 | 60 | 0.6954 | 0.4937 | 0.3567 | 0.1370 | 0.3630 | 0.1433 |
0.7041 | 3.2 | 80 | 0.6946 | 0.5051 | 0.2191 | 0.2860 | 0.2140 | 0.2809 |
0.6975 | 4.0 | 100 | 0.6947 | 0.5013 | 0.1799 | 0.3213 | 0.1787 | 0.3201 |
0.6996 | 4.8 | 120 | 0.6948 | 0.5025 | 0.1521 | 0.3504 | 0.1496 | 0.3479 |
0.7008 | 5.6 | 140 | 0.6944 | 0.4975 | 0.2841 | 0.2134 | 0.2866 | 0.2159 |
0.7004 | 6.4 | 160 | 0.6943 | 0.4968 | 0.1850 | 0.3119 | 0.1881 | 0.3150 |
0.6913 | 7.2 | 180 | 0.6944 | 0.4924 | 0.1553 | 0.3371 | 0.1629 | 0.3447 |
0.703 | 8.0 | 200 | 0.6941 | 0.5025 | 0.2784 | 0.2241 | 0.2759 | 0.2216 |
0.6975 | 8.8 | 220 | 0.6941 | 0.4987 | 0.2197 | 0.2790 | 0.2210 | 0.2803 |
0.6964 | 9.6 | 240 | 0.6942 | 0.4949 | 0.2058 | 0.2891 | 0.2109 | 0.2942 |
0.692 | 10.4 | 260 | 0.6943 | 0.4949 | 0.3037 | 0.1913 | 0.3087 | 0.1963 |
0.6939 | 11.2 | 280 | 0.6943 | 0.4987 | 0.1900 | 0.3087 | 0.1913 | 0.3100 |
0.7043 | 12.0 | 300 | 0.6942 | 0.5044 | 0.2551 | 0.2494 | 0.2506 | 0.2449 |
0.7036 | 12.8 | 320 | 0.6942 | 0.4912 | 0.2102 | 0.2809 | 0.2191 | 0.2898 |
0.697 | 13.6 | 340 | 0.6943 | 0.4975 | 0.1604 | 0.3371 | 0.1629 | 0.3396 |
0.7028 | 14.4 | 360 | 0.6950 | 0.5032 | 0.3939 | 0.1092 | 0.3908 | 0.1061 |
0.7012 | 15.2 | 380 | 0.6940 | 0.4962 | 0.2045 | 0.2917 | 0.2083 | 0.2955 |
0.6976 | 16.0 | 400 | 0.6940 | 0.4968 | 0.2102 | 0.2866 | 0.2134 | 0.2898 |
0.695 | 16.8 | 420 | 0.6944 | 0.5095 | 0.1452 | 0.3643 | 0.1357 | 0.3548 |
0.6985 | 17.6 | 440 | 0.6939 | 0.5013 | 0.2210 | 0.2803 | 0.2197 | 0.2790 |
0.6946 | 18.4 | 460 | 0.6939 | 0.5032 | 0.2765 | 0.2266 | 0.2734 | 0.2235 |
0.6975 | 19.2 | 480 | 0.6940 | 0.4962 | 0.1749 | 0.3213 | 0.1787 | 0.3251 |
0.6958 | 20.0 | 500 | 0.6939 | 0.4905 | 0.2058 | 0.2847 | 0.2153 | 0.2942 |
0.6947 | 20.8 | 520 | 0.6938 | 0.5057 | 0.2771 | 0.2285 | 0.2715 | 0.2229 |
0.7044 | 21.6 | 540 | 0.6940 | 0.5019 | 0.2986 | 0.2033 | 0.2967 | 0.2014 |
0.698 | 22.4 | 560 | 0.6941 | 0.4918 | 0.3201 | 0.1717 | 0.3283 | 0.1799 |
0.7016 | 23.2 | 580 | 0.6939 | 0.5076 | 0.2771 | 0.2304 | 0.2696 | 0.2229 |
0.7029 | 24.0 | 600 | 0.6939 | 0.5063 | 0.2765 | 0.2298 | 0.2702 | 0.2235 |
0.6975 | 24.8 | 620 | 0.6938 | 0.5025 | 0.2904 | 0.2121 | 0.2879 | 0.2096 |
0.6966 | 25.6 | 640 | 0.6940 | 0.5032 | 0.1660 | 0.3371 | 0.1629 | 0.3340 |
0.6974 | 26.4 | 660 | 0.6938 | 0.4994 | 0.1926 | 0.3068 | 0.1932 | 0.3074 |
0.6998 | 27.2 | 680 | 0.6938 | 0.5013 | 0.2229 | 0.2784 | 0.2216 | 0.2771 |
0.6899 | 28.0 | 700 | 0.6937 | 0.5082 | 0.25 | 0.2582 | 0.2418 | 0.25 |
0.6954 | 28.8 | 720 | 0.6937 | 0.4968 | 0.2109 | 0.2860 | 0.2140 | 0.2891 |
0.6926 | 29.6 | 740 | 0.6941 | 0.4899 | 0.3479 | 0.1420 | 0.3580 | 0.1521 |
0.6936 | 30.4 | 760 | 0.6938 | 0.5006 | 0.2822 | 0.2184 | 0.2816 | 0.2178 |
0.6911 | 31.2 | 780 | 0.6937 | 0.5057 | 0.2519 | 0.2538 | 0.2462 | 0.2481 |
0.69 | 32.0 | 800 | 0.6938 | 0.5038 | 0.2904 | 0.2134 | 0.2866 | 0.2096 |
0.6953 | 32.8 | 820 | 0.6937 | 0.5051 | 0.2765 | 0.2285 | 0.2715 | 0.2235 |
0.6971 | 33.6 | 840 | 0.6937 | 0.4956 | 0.2020 | 0.2936 | 0.2064 | 0.2980 |
0.6983 | 34.4 | 860 | 0.6937 | 0.5025 | 0.2727 | 0.2298 | 0.2702 | 0.2273 |
0.698 | 35.2 | 880 | 0.6938 | 0.4987 | 0.3024 | 0.1963 | 0.3037 | 0.1976 |
0.6949 | 36.0 | 900 | 0.6938 | 0.5032 | 0.3081 | 0.1951 | 0.3049 | 0.1919 |
0.6969 | 36.8 | 920 | 0.6937 | 0.5082 | 0.2885 | 0.2197 | 0.2803 | 0.2115 |
0.6978 | 37.6 | 940 | 0.6937 | 0.5088 | 0.3087 | 0.2001 | 0.2999 | 0.1913 |
0.6965 | 38.4 | 960 | 0.6936 | 0.5088 | 0.2588 | 0.25 | 0.25 | 0.2412 |
0.6929 | 39.2 | 980 | 0.6936 | 0.5101 | 0.2620 | 0.2481 | 0.2519 | 0.2380 |
0.6967 | 40.0 | 1000 | 0.6936 | 0.5101 | 0.2702 | 0.2399 | 0.2601 | 0.2298 |
0.6971 | 40.8 | 1020 | 0.6936 | 0.5069 | 0.2431 | 0.2639 | 0.2361 | 0.2569 |
0.6976 | 41.6 | 1040 | 0.6936 | 0.5063 | 0.2418 | 0.2645 | 0.2355 | 0.2582 |
0.6989 | 42.4 | 1060 | 0.6936 | 0.5038 | 0.2304 | 0.2734 | 0.2266 | 0.2696 |
0.6995 | 43.2 | 1080 | 0.6936 | 0.5019 | 0.2254 | 0.2765 | 0.2235 | 0.2746 |
0.6981 | 44.0 | 1100 | 0.6936 | 0.5069 | 0.2386 | 0.2683 | 0.2317 | 0.2614 |
0.6914 | 44.8 | 1120 | 0.6936 | 0.5095 | 0.25 | 0.2595 | 0.2405 | 0.25 |
0.6936 | 45.6 | 1140 | 0.6936 | 0.5095 | 0.25 | 0.2595 | 0.2405 | 0.25 |
0.6951 | 46.4 | 1160 | 0.6936 | 0.5107 | 0.2734 | 0.2374 | 0.2626 | 0.2266 |
0.6964 | 47.2 | 1180 | 0.6936 | 0.5114 | 0.2854 | 0.2260 | 0.2740 | 0.2146 |
0.7004 | 48.0 | 1200 | 0.6936 | 0.5114 | 0.2822 | 0.2292 | 0.2708 | 0.2178 |
0.696 | 48.8 | 1220 | 0.6936 | 0.5088 | 0.2759 | 0.2330 | 0.2670 | 0.2241 |
0.6966 | 49.6 | 1240 | 0.6936 | 0.5114 | 0.2734 | 0.2380 | 0.2620 | 0.2266 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1
- Datasets 2.10.1
- Tokenizers 0.13.2