<!-- 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
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: 0.8584
- Accuracy: 0.8
- Brier Loss: 0.3083
- Nll: 1.3299
- F1 Micro: 0.8000
- F1 Macro: 0.7728
- Ece: 0.2079
- Aurc: 0.0851
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 | 0.8895 | 0.82 | 0.3092 | 1.1901 | 0.82 | 0.8049 | 0.2293 | 0.0751 |
No log | 1.96 | 6 | 0.8886 | 0.81 | 0.3071 | 1.1861 | 0.81 | 0.7912 | 0.2245 | 0.0705 |
No log | 2.96 | 9 | 0.8747 | 0.815 | 0.3065 | 1.1876 | 0.815 | 0.8021 | 0.2265 | 0.0734 |
No log | 3.96 | 12 | 0.8812 | 0.805 | 0.3085 | 1.2661 | 0.805 | 0.7783 | 0.2087 | 0.0761 |
No log | 4.96 | 15 | 0.8874 | 0.81 | 0.3080 | 1.1831 | 0.81 | 0.7871 | 0.2325 | 0.0786 |
No log | 5.96 | 18 | 0.8818 | 0.81 | 0.3089 | 1.2715 | 0.81 | 0.7960 | 0.2345 | 0.0788 |
No log | 6.96 | 21 | 0.8790 | 0.81 | 0.3045 | 1.2619 | 0.81 | 0.7904 | 0.2235 | 0.0693 |
No log | 7.96 | 24 | 0.8794 | 0.805 | 0.3084 | 1.2566 | 0.805 | 0.7884 | 0.2205 | 0.0787 |
No log | 8.96 | 27 | 0.8838 | 0.815 | 0.3134 | 1.3380 | 0.815 | 0.8072 | 0.2230 | 0.0751 |
No log | 9.96 | 30 | 0.8849 | 0.8 | 0.3132 | 1.3205 | 0.8000 | 0.7757 | 0.2229 | 0.0824 |
No log | 10.96 | 33 | 0.8633 | 0.81 | 0.3061 | 1.3978 | 0.81 | 0.7938 | 0.2004 | 0.0756 |
No log | 11.96 | 36 | 0.8746 | 0.81 | 0.3089 | 1.3970 | 0.81 | 0.7918 | 0.2346 | 0.0741 |
No log | 12.96 | 39 | 0.8625 | 0.805 | 0.3078 | 1.1961 | 0.805 | 0.7945 | 0.2505 | 0.0854 |
No log | 13.96 | 42 | 0.8636 | 0.815 | 0.3068 | 1.2113 | 0.815 | 0.8046 | 0.2371 | 0.0804 |
No log | 14.96 | 45 | 0.8906 | 0.79 | 0.3157 | 1.3748 | 0.79 | 0.7777 | 0.2279 | 0.0847 |
No log | 15.96 | 48 | 0.8601 | 0.805 | 0.3040 | 1.2977 | 0.805 | 0.7876 | 0.2176 | 0.0805 |
No log | 16.96 | 51 | 0.8606 | 0.815 | 0.3083 | 1.4136 | 0.815 | 0.8077 | 0.2279 | 0.0787 |
No log | 17.96 | 54 | 0.9013 | 0.8 | 0.3261 | 1.2494 | 0.8000 | 0.7886 | 0.2194 | 0.0871 |
No log | 18.96 | 57 | 0.8653 | 0.805 | 0.3143 | 1.4166 | 0.805 | 0.7935 | 0.2170 | 0.0786 |
No log | 19.96 | 60 | 0.8459 | 0.81 | 0.3030 | 1.2629 | 0.81 | 0.7953 | 0.2129 | 0.0892 |
No log | 20.96 | 63 | 0.8689 | 0.795 | 0.3106 | 1.2823 | 0.795 | 0.7725 | 0.2099 | 0.0828 |
No log | 21.96 | 66 | 0.8563 | 0.81 | 0.3016 | 1.2789 | 0.81 | 0.7954 | 0.2324 | 0.0742 |
No log | 22.96 | 69 | 0.8998 | 0.785 | 0.3231 | 1.6511 | 0.785 | 0.7642 | 0.2178 | 0.1015 |
No log | 23.96 | 72 | 0.8338 | 0.805 | 0.2971 | 1.0504 | 0.805 | 0.7868 | 0.2135 | 0.0645 |
No log | 24.96 | 75 | 0.8423 | 0.8 | 0.3040 | 1.4777 | 0.8000 | 0.7771 | 0.2283 | 0.0689 |
No log | 25.96 | 78 | 0.8775 | 0.8 | 0.3218 | 1.4206 | 0.8000 | 0.7774 | 0.2204 | 0.1120 |
No log | 26.96 | 81 | 0.8389 | 0.8 | 0.2984 | 1.1946 | 0.8000 | 0.7771 | 0.1990 | 0.0737 |
No log | 27.96 | 84 | 0.9119 | 0.795 | 0.3319 | 1.6978 | 0.795 | 0.7805 | 0.2279 | 0.1109 |
No log | 28.96 | 87 | 0.8689 | 0.805 | 0.3144 | 1.2644 | 0.805 | 0.7971 | 0.2216 | 0.0787 |
No log | 29.96 | 90 | 0.8404 | 0.8 | 0.2990 | 1.1775 | 0.8000 | 0.7848 | 0.1962 | 0.0805 |
No log | 30.96 | 93 | 0.8842 | 0.8 | 0.3226 | 1.3091 | 0.8000 | 0.7904 | 0.2168 | 0.1020 |
No log | 31.96 | 96 | 0.8653 | 0.805 | 0.3086 | 1.3926 | 0.805 | 0.7818 | 0.1996 | 0.0853 |
No log | 32.96 | 99 | 0.8767 | 0.785 | 0.3142 | 1.2268 | 0.785 | 0.7684 | 0.2117 | 0.0739 |
No log | 33.96 | 102 | 0.9349 | 0.775 | 0.3410 | 1.3988 | 0.775 | 0.7600 | 0.2246 | 0.1024 |
No log | 34.96 | 105 | 0.8606 | 0.79 | 0.3035 | 1.0902 | 0.79 | 0.7683 | 0.1954 | 0.0830 |
No log | 35.96 | 108 | 0.8578 | 0.815 | 0.3050 | 1.3418 | 0.815 | 0.7923 | 0.2155 | 0.0923 |
No log | 36.96 | 111 | 0.8641 | 0.795 | 0.3128 | 1.2449 | 0.795 | 0.7694 | 0.2068 | 0.0878 |
No log | 37.96 | 114 | 0.8489 | 0.8 | 0.2996 | 1.2505 | 0.8000 | 0.7698 | 0.2027 | 0.0827 |
No log | 38.96 | 117 | 0.8465 | 0.82 | 0.3011 | 1.3264 | 0.82 | 0.7947 | 0.2033 | 0.0923 |
No log | 39.96 | 120 | 0.8608 | 0.8 | 0.3051 | 1.3178 | 0.8000 | 0.7706 | 0.2072 | 0.0894 |
No log | 40.96 | 123 | 0.8592 | 0.8 | 0.3066 | 1.3141 | 0.8000 | 0.7692 | 0.2069 | 0.0909 |
No log | 41.96 | 126 | 0.8611 | 0.805 | 0.3125 | 1.2988 | 0.805 | 0.7832 | 0.2094 | 0.0791 |
No log | 42.96 | 129 | 0.8516 | 0.805 | 0.3000 | 1.3221 | 0.805 | 0.7791 | 0.2179 | 0.0884 |
No log | 43.96 | 132 | 0.8587 | 0.8 | 0.3064 | 1.3414 | 0.8000 | 0.7784 | 0.2056 | 0.0922 |
No log | 44.96 | 135 | 0.8691 | 0.79 | 0.3181 | 1.3262 | 0.79 | 0.7765 | 0.2153 | 0.0884 |
No log | 45.96 | 138 | 0.8576 | 0.81 | 0.3066 | 1.1918 | 0.81 | 0.7847 | 0.2182 | 0.1009 |
No log | 46.96 | 141 | 0.8722 | 0.8 | 0.3152 | 1.4909 | 0.8000 | 0.7798 | 0.2219 | 0.1012 |
No log | 47.96 | 144 | 0.8399 | 0.81 | 0.3087 | 1.5338 | 0.81 | 0.7849 | 0.2138 | 0.0740 |
No log | 48.96 | 147 | 0.8393 | 0.805 | 0.3004 | 1.3810 | 0.805 | 0.7819 | 0.2150 | 0.0696 |
No log | 49.96 | 150 | 0.8899 | 0.78 | 0.3201 | 1.5622 | 0.78 | 0.7644 | 0.2227 | 0.0960 |
No log | 50.96 | 153 | 0.8954 | 0.78 | 0.3249 | 1.6494 | 0.78 | 0.7654 | 0.2135 | 0.0902 |
No log | 51.96 | 156 | 0.8259 | 0.79 | 0.2954 | 1.2271 | 0.79 | 0.7707 | 0.2129 | 0.0659 |
No log | 52.96 | 159 | 0.8806 | 0.795 | 0.3145 | 1.4079 | 0.795 | 0.7759 | 0.2046 | 0.0877 |
No log | 53.96 | 162 | 0.8842 | 0.81 | 0.3178 | 1.3465 | 0.81 | 0.7925 | 0.2173 | 0.1037 |
No log | 54.96 | 165 | 0.8741 | 0.8 | 0.3173 | 1.4540 | 0.8000 | 0.7750 | 0.2079 | 0.0819 |
No log | 55.96 | 168 | 0.8242 | 0.8 | 0.2964 | 1.3053 | 0.8000 | 0.7838 | 0.1972 | 0.0670 |
No log | 56.96 | 171 | 0.8350 | 0.825 | 0.2962 | 1.2110 | 0.825 | 0.8135 | 0.2126 | 0.0780 |
No log | 57.96 | 174 | 0.8491 | 0.815 | 0.3034 | 1.3250 | 0.815 | 0.8070 | 0.2116 | 0.0875 |
No log | 58.96 | 177 | 0.8584 | 0.795 | 0.3119 | 1.3162 | 0.795 | 0.7764 | 0.1956 | 0.0860 |
No log | 59.96 | 180 | 0.8546 | 0.79 | 0.3115 | 1.3315 | 0.79 | 0.7740 | 0.1855 | 0.0828 |
No log | 60.96 | 183 | 0.8564 | 0.79 | 0.3068 | 1.3275 | 0.79 | 0.7760 | 0.2008 | 0.0862 |
No log | 61.96 | 186 | 0.8573 | 0.795 | 0.3068 | 1.3160 | 0.795 | 0.7738 | 0.2117 | 0.0884 |
No log | 62.96 | 189 | 0.8503 | 0.785 | 0.3088 | 1.3498 | 0.785 | 0.7650 | 0.2069 | 0.0856 |
No log | 63.96 | 192 | 0.8639 | 0.81 | 0.3111 | 1.2614 | 0.81 | 0.7873 | 0.2247 | 0.0893 |
No log | 64.96 | 195 | 0.8744 | 0.805 | 0.3128 | 1.3294 | 0.805 | 0.7888 | 0.2096 | 0.0912 |
No log | 65.96 | 198 | 0.8727 | 0.8 | 0.3138 | 1.4212 | 0.8000 | 0.7903 | 0.2031 | 0.0849 |
No log | 66.96 | 201 | 0.8612 | 0.79 | 0.3084 | 1.3592 | 0.79 | 0.7702 | 0.1855 | 0.0816 |
No log | 67.96 | 204 | 0.8576 | 0.79 | 0.3071 | 1.4005 | 0.79 | 0.7667 | 0.1896 | 0.0863 |
No log | 68.96 | 207 | 0.8540 | 0.805 | 0.3037 | 1.3957 | 0.805 | 0.7775 | 0.2263 | 0.0876 |
No log | 69.96 | 210 | 0.8499 | 0.81 | 0.2982 | 1.3987 | 0.81 | 0.7874 | 0.2109 | 0.0856 |
No log | 70.96 | 213 | 0.8465 | 0.815 | 0.3001 | 1.3222 | 0.815 | 0.7901 | 0.2224 | 0.0928 |
No log | 71.96 | 216 | 0.8541 | 0.81 | 0.3041 | 1.3331 | 0.81 | 0.7827 | 0.2169 | 0.0897 |
No log | 72.96 | 219 | 0.8546 | 0.795 | 0.3066 | 1.3991 | 0.795 | 0.7720 | 0.2141 | 0.0871 |
No log | 73.96 | 222 | 0.8569 | 0.79 | 0.3039 | 1.3544 | 0.79 | 0.7672 | 0.1958 | 0.0863 |
No log | 74.96 | 225 | 0.8622 | 0.805 | 0.3028 | 1.3384 | 0.805 | 0.7847 | 0.1938 | 0.0879 |
No log | 75.96 | 228 | 0.8610 | 0.805 | 0.3039 | 1.3285 | 0.805 | 0.7810 | 0.2033 | 0.0947 |
No log | 76.96 | 231 | 0.8581 | 0.81 | 0.3031 | 1.3334 | 0.81 | 0.7840 | 0.1993 | 0.0944 |
No log | 77.96 | 234 | 0.8607 | 0.8 | 0.3055 | 1.3260 | 0.8000 | 0.7785 | 0.1979 | 0.0899 |
No log | 78.96 | 237 | 0.8642 | 0.79 | 0.3068 | 1.3928 | 0.79 | 0.7672 | 0.1822 | 0.0869 |
No log | 79.96 | 240 | 0.8640 | 0.805 | 0.3044 | 1.3311 | 0.805 | 0.7786 | 0.2001 | 0.0916 |
No log | 80.96 | 243 | 0.8648 | 0.81 | 0.3056 | 1.2812 | 0.81 | 0.7836 | 0.2173 | 0.0955 |
No log | 81.96 | 246 | 0.8639 | 0.825 | 0.3056 | 1.3295 | 0.825 | 0.8062 | 0.1952 | 0.0913 |
No log | 82.96 | 249 | 0.8643 | 0.805 | 0.3082 | 1.3334 | 0.805 | 0.7887 | 0.2108 | 0.0881 |
No log | 83.96 | 252 | 0.8626 | 0.795 | 0.3068 | 1.3334 | 0.795 | 0.7780 | 0.2097 | 0.0845 |
No log | 84.96 | 255 | 0.8586 | 0.81 | 0.3033 | 1.2646 | 0.81 | 0.7893 | 0.2035 | 0.0808 |
No log | 85.96 | 258 | 0.8570 | 0.805 | 0.3024 | 1.2694 | 0.805 | 0.7802 | 0.1947 | 0.0811 |
No log | 86.96 | 261 | 0.8557 | 0.795 | 0.3023 | 1.3261 | 0.795 | 0.7657 | 0.1966 | 0.0828 |
No log | 87.96 | 264 | 0.8576 | 0.8 | 0.3051 | 1.3283 | 0.8000 | 0.7754 | 0.2072 | 0.0848 |
No log | 88.96 | 267 | 0.8537 | 0.8 | 0.3083 | 1.3257 | 0.8000 | 0.7771 | 0.2167 | 0.0859 |
No log | 89.96 | 270 | 0.8591 | 0.795 | 0.3106 | 1.3262 | 0.795 | 0.7737 | 0.2011 | 0.0866 |
No log | 90.96 | 273 | 0.8612 | 0.785 | 0.3122 | 1.3279 | 0.785 | 0.7594 | 0.1885 | 0.0868 |
No log | 91.96 | 276 | 0.8571 | 0.795 | 0.3104 | 1.3248 | 0.795 | 0.7667 | 0.1966 | 0.0853 |
No log | 92.96 | 279 | 0.8560 | 0.795 | 0.3082 | 1.3244 | 0.795 | 0.7667 | 0.2147 | 0.0836 |
No log | 93.96 | 282 | 0.8551 | 0.8 | 0.3071 | 1.3251 | 0.8000 | 0.7766 | 0.2109 | 0.0830 |
No log | 94.96 | 285 | 0.8556 | 0.79 | 0.3076 | 1.3264 | 0.79 | 0.7577 | 0.1885 | 0.0834 |
No log | 95.96 | 288 | 0.8569 | 0.795 | 0.3078 | 1.3280 | 0.795 | 0.7675 | 0.1980 | 0.0840 |
No log | 96.96 | 291 | 0.8581 | 0.795 | 0.3082 | 1.3290 | 0.795 | 0.7675 | 0.2039 | 0.0842 |
No log | 97.96 | 294 | 0.8585 | 0.8 | 0.3084 | 1.3300 | 0.8000 | 0.7728 | 0.2137 | 0.0849 |
No log | 98.96 | 297 | 0.8589 | 0.8 | 0.3083 | 1.3301 | 0.8000 | 0.7728 | 0.2156 | 0.0850 |
No log | 99.96 | 300 | 0.8584 | 0.8 | 0.3083 | 1.3299 | 0.8000 | 0.7728 | 0.2079 | 0.0851 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2