<!-- 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. -->
vit-base_tobacco_crl_allv2
This model is a fine-tuned version of jordyvl/vit-base_tobacco on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8070
- Accuracy: 0.82
- Brier Loss: 0.2884
- Nll: 1.3803
- F1 Micro: 0.82
- F1 Macro: 0.8075
- Ece: 0.1968
- Aurc: 0.0591
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 100
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Brier Loss | Nll | F1 Micro | F1 Macro | Ece | Aurc |
---|---|---|---|---|---|---|---|---|---|---|
No log | 0.96 | 3 | 0.8264 | 0.815 | 0.3085 | 1.1929 | 0.815 | 0.7922 | 0.2249 | 0.0764 |
No log | 1.96 | 6 | 0.8371 | 0.815 | 0.3061 | 1.1842 | 0.815 | 0.7910 | 0.2299 | 0.0718 |
No log | 2.96 | 9 | 0.8530 | 0.8 | 0.3072 | 1.1942 | 0.8000 | 0.7798 | 0.2335 | 0.0749 |
No log | 3.96 | 12 | 0.8469 | 0.82 | 0.3056 | 1.1930 | 0.82 | 0.8049 | 0.2385 | 0.0743 |
No log | 4.96 | 15 | 0.8274 | 0.84 | 0.3006 | 1.3642 | 0.8400 | 0.8268 | 0.2376 | 0.0716 |
No log | 5.96 | 18 | 0.8632 | 0.81 | 0.3105 | 1.2012 | 0.81 | 0.7873 | 0.2468 | 0.0806 |
No log | 6.96 | 21 | 0.8216 | 0.815 | 0.2949 | 1.1817 | 0.815 | 0.8037 | 0.2343 | 0.0623 |
No log | 7.96 | 24 | 0.8540 | 0.805 | 0.3056 | 1.1903 | 0.805 | 0.7842 | 0.1968 | 0.0782 |
No log | 8.96 | 27 | 0.8223 | 0.82 | 0.2955 | 1.3001 | 0.82 | 0.8114 | 0.2195 | 0.0629 |
No log | 9.96 | 30 | 0.8432 | 0.81 | 0.3025 | 1.1862 | 0.81 | 0.7999 | 0.2074 | 0.0752 |
No log | 10.96 | 33 | 0.8445 | 0.795 | 0.3049 | 1.2622 | 0.795 | 0.7820 | 0.2254 | 0.0629 |
No log | 11.96 | 36 | 0.8367 | 0.81 | 0.2987 | 1.2088 | 0.81 | 0.7945 | 0.2269 | 0.0736 |
No log | 12.96 | 39 | 0.8365 | 0.81 | 0.3022 | 1.4703 | 0.81 | 0.7926 | 0.2216 | 0.0733 |
No log | 13.96 | 42 | 0.8460 | 0.805 | 0.3020 | 1.2887 | 0.805 | 0.7864 | 0.2151 | 0.0673 |
No log | 14.96 | 45 | 0.8791 | 0.805 | 0.3161 | 1.2981 | 0.805 | 0.7822 | 0.2089 | 0.0840 |
No log | 15.96 | 48 | 0.8318 | 0.805 | 0.3049 | 1.3976 | 0.805 | 0.7900 | 0.2216 | 0.0787 |
No log | 16.96 | 51 | 0.9041 | 0.795 | 0.3295 | 1.3060 | 0.795 | 0.7798 | 0.2362 | 0.0772 |
No log | 17.96 | 54 | 0.8434 | 0.81 | 0.3051 | 1.4598 | 0.81 | 0.7863 | 0.2121 | 0.0756 |
No log | 18.96 | 57 | 0.8693 | 0.8 | 0.3080 | 1.4271 | 0.8000 | 0.7818 | 0.2227 | 0.0780 |
No log | 19.96 | 60 | 0.8635 | 0.795 | 0.3138 | 1.4281 | 0.795 | 0.7611 | 0.2259 | 0.0764 |
No log | 20.96 | 63 | 0.9061 | 0.78 | 0.3285 | 1.3005 | 0.78 | 0.7667 | 0.2245 | 0.0788 |
No log | 21.96 | 66 | 0.8564 | 0.805 | 0.3065 | 1.3226 | 0.805 | 0.7827 | 0.2093 | 0.0901 |
No log | 22.96 | 69 | 0.8354 | 0.78 | 0.3035 | 1.2478 | 0.78 | 0.7556 | 0.1985 | 0.0633 |
No log | 23.96 | 72 | 0.8458 | 0.8 | 0.3068 | 1.1995 | 0.8000 | 0.7869 | 0.2117 | 0.0666 |
No log | 24.96 | 75 | 0.8406 | 0.78 | 0.3055 | 1.1192 | 0.78 | 0.7605 | 0.2131 | 0.0646 |
No log | 25.96 | 78 | 0.8048 | 0.81 | 0.2904 | 1.1867 | 0.81 | 0.7937 | 0.1940 | 0.0611 |
No log | 26.96 | 81 | 0.8429 | 0.815 | 0.3050 | 1.4600 | 0.815 | 0.7924 | 0.2020 | 0.0801 |
No log | 27.96 | 84 | 0.8362 | 0.785 | 0.3042 | 1.4610 | 0.785 | 0.7667 | 0.1953 | 0.0733 |
No log | 28.96 | 87 | 0.8709 | 0.795 | 0.3096 | 1.5073 | 0.795 | 0.7787 | 0.2015 | 0.0844 |
No log | 29.96 | 90 | 0.8696 | 0.785 | 0.3223 | 1.4348 | 0.785 | 0.7607 | 0.2249 | 0.0798 |
No log | 30.96 | 93 | 0.8748 | 0.795 | 0.3133 | 1.5627 | 0.795 | 0.7759 | 0.2125 | 0.0919 |
No log | 31.96 | 96 | 0.8320 | 0.815 | 0.3006 | 1.4744 | 0.815 | 0.8039 | 0.2239 | 0.0816 |
No log | 32.96 | 99 | 0.8299 | 0.79 | 0.3041 | 1.3752 | 0.79 | 0.7725 | 0.2193 | 0.0726 |
No log | 33.96 | 102 | 0.8377 | 0.795 | 0.3058 | 1.3802 | 0.795 | 0.7769 | 0.2196 | 0.0773 |
No log | 34.96 | 105 | 0.8195 | 0.8 | 0.2986 | 1.4524 | 0.8000 | 0.7836 | 0.2032 | 0.0756 |
No log | 35.96 | 108 | 0.8600 | 0.785 | 0.3081 | 1.4267 | 0.785 | 0.7599 | 0.2042 | 0.0914 |
No log | 36.96 | 111 | 0.8565 | 0.795 | 0.3073 | 1.3142 | 0.795 | 0.7720 | 0.2018 | 0.0773 |
No log | 37.96 | 114 | 0.8175 | 0.815 | 0.3002 | 1.3310 | 0.815 | 0.8107 | 0.2289 | 0.0659 |
No log | 38.96 | 117 | 0.8309 | 0.825 | 0.2975 | 1.3417 | 0.825 | 0.8100 | 0.1944 | 0.0890 |
No log | 39.96 | 120 | 0.8479 | 0.795 | 0.3071 | 1.3669 | 0.795 | 0.7796 | 0.2001 | 0.0796 |
No log | 40.96 | 123 | 0.8052 | 0.825 | 0.2940 | 1.4031 | 0.825 | 0.8087 | 0.1969 | 0.0661 |
No log | 41.96 | 126 | 0.8009 | 0.8 | 0.2937 | 1.1756 | 0.8000 | 0.7867 | 0.2003 | 0.0624 |
No log | 42.96 | 129 | 0.8367 | 0.79 | 0.3059 | 1.2063 | 0.79 | 0.7746 | 0.1964 | 0.0657 |
No log | 43.96 | 132 | 0.8222 | 0.81 | 0.3021 | 1.2555 | 0.81 | 0.7994 | 0.2268 | 0.0695 |
No log | 44.96 | 135 | 0.8234 | 0.805 | 0.2990 | 1.2438 | 0.805 | 0.7889 | 0.2045 | 0.0820 |
No log | 45.96 | 138 | 0.8378 | 0.81 | 0.3096 | 1.3797 | 0.81 | 0.7904 | 0.2136 | 0.0761 |
No log | 46.96 | 141 | 0.8089 | 0.8 | 0.2959 | 1.2355 | 0.8000 | 0.7837 | 0.1987 | 0.0716 |
No log | 47.96 | 144 | 0.8427 | 0.79 | 0.3052 | 1.3295 | 0.79 | 0.7746 | 0.2032 | 0.0794 |
No log | 48.96 | 147 | 0.8269 | 0.81 | 0.3034 | 1.3293 | 0.81 | 0.7968 | 0.2014 | 0.0826 |
No log | 49.96 | 150 | 0.8081 | 0.81 | 0.2958 | 1.3146 | 0.81 | 0.7901 | 0.2004 | 0.0721 |
No log | 50.96 | 153 | 0.8084 | 0.8 | 0.2967 | 1.3800 | 0.8000 | 0.7799 | 0.2114 | 0.0623 |
No log | 51.96 | 156 | 0.8076 | 0.805 | 0.2931 | 1.3180 | 0.805 | 0.7850 | 0.2068 | 0.0631 |
No log | 52.96 | 159 | 0.8163 | 0.82 | 0.3025 | 1.3950 | 0.82 | 0.8028 | 0.2294 | 0.0694 |
No log | 53.96 | 162 | 0.8519 | 0.765 | 0.3193 | 1.3976 | 0.765 | 0.7522 | 0.2062 | 0.0703 |
No log | 54.96 | 165 | 0.8146 | 0.79 | 0.2991 | 1.3098 | 0.79 | 0.7738 | 0.1900 | 0.0615 |
No log | 55.96 | 168 | 0.8064 | 0.815 | 0.2918 | 1.3889 | 0.815 | 0.7978 | 0.2099 | 0.0719 |
No log | 56.96 | 171 | 0.8225 | 0.81 | 0.2991 | 1.3051 | 0.81 | 0.7916 | 0.2153 | 0.0794 |
No log | 57.96 | 174 | 0.8222 | 0.815 | 0.3021 | 1.2774 | 0.815 | 0.8011 | 0.2052 | 0.0722 |
No log | 58.96 | 177 | 0.8025 | 0.82 | 0.2949 | 1.4508 | 0.82 | 0.8103 | 0.2228 | 0.0636 |
No log | 59.96 | 180 | 0.7993 | 0.8 | 0.2911 | 1.3141 | 0.8000 | 0.7829 | 0.2008 | 0.0685 |
No log | 60.96 | 183 | 0.7919 | 0.805 | 0.2900 | 1.3050 | 0.805 | 0.7891 | 0.1983 | 0.0618 |
No log | 61.96 | 186 | 0.7994 | 0.81 | 0.2921 | 1.2423 | 0.81 | 0.7964 | 0.1920 | 0.0710 |
No log | 62.96 | 189 | 0.7966 | 0.815 | 0.2887 | 1.2438 | 0.815 | 0.7970 | 0.1940 | 0.0727 |
No log | 63.96 | 192 | 0.7892 | 0.82 | 0.2860 | 1.3182 | 0.82 | 0.8032 | 0.2006 | 0.0622 |
No log | 64.96 | 195 | 0.7916 | 0.815 | 0.2880 | 1.2500 | 0.815 | 0.7998 | 0.2005 | 0.0611 |
No log | 65.96 | 198 | 0.7920 | 0.81 | 0.2892 | 1.2549 | 0.81 | 0.7925 | 0.2207 | 0.0595 |
No log | 66.96 | 201 | 0.7924 | 0.81 | 0.2895 | 1.3216 | 0.81 | 0.7925 | 0.2084 | 0.0587 |
No log | 67.96 | 204 | 0.7935 | 0.8 | 0.2898 | 1.2552 | 0.8000 | 0.7858 | 0.2142 | 0.0592 |
No log | 68.96 | 207 | 0.7929 | 0.805 | 0.2894 | 1.3132 | 0.805 | 0.7891 | 0.2224 | 0.0581 |
No log | 69.96 | 210 | 0.7960 | 0.805 | 0.2892 | 1.2494 | 0.805 | 0.7887 | 0.2193 | 0.0606 |
No log | 70.96 | 213 | 0.7928 | 0.815 | 0.2876 | 1.3287 | 0.815 | 0.8024 | 0.2152 | 0.0595 |
No log | 71.96 | 216 | 0.7939 | 0.81 | 0.2875 | 1.2485 | 0.81 | 0.7920 | 0.2085 | 0.0604 |
No log | 72.96 | 219 | 0.7937 | 0.81 | 0.2871 | 1.2538 | 0.81 | 0.7920 | 0.2009 | 0.0600 |
No log | 73.96 | 222 | 0.7924 | 0.82 | 0.2869 | 1.3766 | 0.82 | 0.8095 | 0.1965 | 0.0586 |
No log | 74.96 | 225 | 0.7980 | 0.81 | 0.2884 | 1.2557 | 0.81 | 0.7920 | 0.2197 | 0.0614 |
No log | 75.96 | 228 | 0.7938 | 0.82 | 0.2863 | 1.3205 | 0.82 | 0.8052 | 0.1892 | 0.0598 |
No log | 76.96 | 231 | 0.7935 | 0.825 | 0.2860 | 1.3242 | 0.825 | 0.8123 | 0.1866 | 0.0592 |
No log | 77.96 | 234 | 0.7942 | 0.815 | 0.2864 | 1.3761 | 0.815 | 0.7990 | 0.1821 | 0.0581 |
No log | 78.96 | 237 | 0.7939 | 0.815 | 0.2864 | 1.3762 | 0.815 | 0.7990 | 0.1893 | 0.0578 |
No log | 79.96 | 240 | 0.7961 | 0.815 | 0.2874 | 1.3788 | 0.815 | 0.7990 | 0.1992 | 0.0581 |
No log | 80.96 | 243 | 0.7981 | 0.815 | 0.2880 | 1.3816 | 0.815 | 0.7990 | 0.1946 | 0.0583 |
No log | 81.96 | 246 | 0.7989 | 0.815 | 0.2882 | 1.3820 | 0.815 | 0.7990 | 0.1939 | 0.0584 |
No log | 82.96 | 249 | 0.7995 | 0.81 | 0.2882 | 1.3813 | 0.81 | 0.7957 | 0.1929 | 0.0587 |
No log | 83.96 | 252 | 0.8014 | 0.81 | 0.2886 | 1.3263 | 0.81 | 0.7957 | 0.2109 | 0.0587 |
No log | 84.96 | 255 | 0.7998 | 0.82 | 0.2878 | 1.3793 | 0.82 | 0.8090 | 0.1873 | 0.0582 |
No log | 85.96 | 258 | 0.8021 | 0.82 | 0.2881 | 1.3816 | 0.82 | 0.8090 | 0.2131 | 0.0587 |
No log | 86.96 | 261 | 0.8036 | 0.815 | 0.2884 | 1.3827 | 0.815 | 0.7975 | 0.2065 | 0.0589 |
No log | 87.96 | 264 | 0.8051 | 0.815 | 0.2887 | 1.3836 | 0.815 | 0.7975 | 0.1989 | 0.0594 |
No log | 88.96 | 267 | 0.8053 | 0.815 | 0.2885 | 1.3830 | 0.815 | 0.8034 | 0.1883 | 0.0592 |
No log | 89.96 | 270 | 0.8056 | 0.815 | 0.2883 | 1.3830 | 0.815 | 0.8034 | 0.1977 | 0.0590 |
No log | 90.96 | 273 | 0.8051 | 0.82 | 0.2878 | 1.3815 | 0.82 | 0.8075 | 0.2024 | 0.0586 |
No log | 91.96 | 276 | 0.8052 | 0.82 | 0.2877 | 1.3805 | 0.82 | 0.8075 | 0.2133 | 0.0584 |
No log | 92.96 | 279 | 0.8055 | 0.82 | 0.2879 | 1.3799 | 0.82 | 0.8075 | 0.1953 | 0.0587 |
No log | 93.96 | 282 | 0.8055 | 0.82 | 0.2878 | 1.3793 | 0.82 | 0.8075 | 0.1995 | 0.0586 |
No log | 94.96 | 285 | 0.8061 | 0.815 | 0.2881 | 1.3797 | 0.815 | 0.7975 | 0.1905 | 0.0590 |
No log | 95.96 | 288 | 0.8064 | 0.815 | 0.2882 | 1.3798 | 0.815 | 0.7975 | 0.1990 | 0.0590 |
No log | 96.96 | 291 | 0.8065 | 0.82 | 0.2883 | 1.3799 | 0.82 | 0.8075 | 0.2140 | 0.0588 |
No log | 97.96 | 294 | 0.8069 | 0.815 | 0.2884 | 1.3802 | 0.815 | 0.7975 | 0.1916 | 0.0591 |
No log | 98.96 | 297 | 0.8070 | 0.82 | 0.2884 | 1.3803 | 0.82 | 0.8075 | 0.1968 | 0.0592 |
No log | 99.96 | 300 | 0.8070 | 0.82 | 0.2884 | 1.3803 | 0.82 | 0.8075 | 0.1968 | 0.0591 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2