<!-- 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. -->
6_e_200-tiny_tobacco3482_kd_CEKD_t2.5_a0.5
This model is a fine-tuned version of WinKawaks/vit-tiny-patch16-224 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4137
- Accuracy: 0.83
- Brier Loss: 0.2631
- Nll: 1.5189
- F1 Micro: 0.83
- F1 Macro: 0.8172
- Ece: 0.2007
- 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: 0.0001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- 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 | 1.0 | 25 | 1.6265 | 0.23 | 0.8647 | 5.1432 | 0.23 | 0.1847 | 0.2751 | 0.7516 |
No log | 2.0 | 50 | 1.0240 | 0.505 | 0.6074 | 2.7425 | 0.505 | 0.3980 | 0.2977 | 0.2705 |
No log | 3.0 | 75 | 0.8130 | 0.655 | 0.4809 | 2.4939 | 0.655 | 0.5584 | 0.2576 | 0.1502 |
No log | 4.0 | 100 | 0.6703 | 0.735 | 0.3867 | 1.3509 | 0.735 | 0.6895 | 0.2334 | 0.1109 |
No log | 5.0 | 125 | 0.6313 | 0.755 | 0.3420 | 1.2521 | 0.755 | 0.7207 | 0.2081 | 0.0789 |
No log | 6.0 | 150 | 0.6598 | 0.76 | 0.3543 | 1.4171 | 0.76 | 0.7103 | 0.2275 | 0.0886 |
No log | 7.0 | 175 | 0.5669 | 0.77 | 0.3368 | 1.5060 | 0.7700 | 0.7351 | 0.2247 | 0.0941 |
No log | 8.0 | 200 | 0.5486 | 0.775 | 0.3004 | 1.1511 | 0.775 | 0.7413 | 0.2252 | 0.0640 |
No log | 9.0 | 225 | 0.5456 | 0.795 | 0.3141 | 1.4663 | 0.795 | 0.7762 | 0.2198 | 0.0889 |
No log | 10.0 | 250 | 0.4954 | 0.82 | 0.2819 | 1.4644 | 0.82 | 0.7981 | 0.2150 | 0.0649 |
No log | 11.0 | 275 | 0.4804 | 0.805 | 0.2866 | 1.3705 | 0.805 | 0.7927 | 0.2078 | 0.0658 |
No log | 12.0 | 300 | 0.5234 | 0.785 | 0.3152 | 1.5290 | 0.785 | 0.7681 | 0.2149 | 0.0637 |
No log | 13.0 | 325 | 0.4701 | 0.815 | 0.2839 | 1.4490 | 0.815 | 0.8010 | 0.2315 | 0.0586 |
No log | 14.0 | 350 | 0.4859 | 0.795 | 0.2807 | 1.1224 | 0.795 | 0.7957 | 0.2170 | 0.0512 |
No log | 15.0 | 375 | 0.5580 | 0.79 | 0.3272 | 1.7539 | 0.79 | 0.7735 | 0.2376 | 0.0708 |
No log | 16.0 | 400 | 0.4918 | 0.8 | 0.2961 | 1.5112 | 0.8000 | 0.7988 | 0.1850 | 0.0568 |
No log | 17.0 | 425 | 0.4442 | 0.8 | 0.2846 | 1.6182 | 0.8000 | 0.7767 | 0.2083 | 0.0712 |
No log | 18.0 | 450 | 0.4460 | 0.82 | 0.2760 | 1.6839 | 0.82 | 0.8027 | 0.2127 | 0.0523 |
No log | 19.0 | 475 | 0.4423 | 0.825 | 0.2676 | 1.3774 | 0.825 | 0.8176 | 0.1853 | 0.0557 |
0.4472 | 20.0 | 500 | 0.4998 | 0.81 | 0.2910 | 1.7711 | 0.81 | 0.8152 | 0.2181 | 0.0635 |
0.4472 | 21.0 | 525 | 0.4579 | 0.83 | 0.2871 | 1.7025 | 0.83 | 0.8135 | 0.1927 | 0.0696 |
0.4472 | 22.0 | 550 | 0.4421 | 0.825 | 0.2683 | 1.6453 | 0.825 | 0.8215 | 0.1929 | 0.0613 |
0.4472 | 23.0 | 575 | 0.4368 | 0.8 | 0.2821 | 1.7298 | 0.8000 | 0.7684 | 0.2060 | 0.0771 |
0.4472 | 24.0 | 600 | 0.4310 | 0.83 | 0.2689 | 1.4699 | 0.83 | 0.8163 | 0.2067 | 0.0556 |
0.4472 | 25.0 | 625 | 0.4394 | 0.83 | 0.2751 | 1.5955 | 0.83 | 0.8166 | 0.2138 | 0.0681 |
0.4472 | 26.0 | 650 | 0.4395 | 0.815 | 0.2786 | 1.6788 | 0.815 | 0.8033 | 0.2034 | 0.0643 |
0.4472 | 27.0 | 675 | 0.4118 | 0.84 | 0.2578 | 1.5641 | 0.8400 | 0.8293 | 0.2024 | 0.0554 |
0.4472 | 28.0 | 700 | 0.4273 | 0.82 | 0.2707 | 1.7118 | 0.82 | 0.8090 | 0.2133 | 0.0674 |
0.4472 | 29.0 | 725 | 0.4207 | 0.835 | 0.2648 | 1.6469 | 0.835 | 0.8206 | 0.1948 | 0.0652 |
0.4472 | 30.0 | 750 | 0.4172 | 0.825 | 0.2620 | 1.5024 | 0.825 | 0.8114 | 0.1833 | 0.0601 |
0.4472 | 31.0 | 775 | 0.4148 | 0.825 | 0.2610 | 1.4994 | 0.825 | 0.8070 | 0.2052 | 0.0593 |
0.4472 | 32.0 | 800 | 0.4148 | 0.825 | 0.2627 | 1.6293 | 0.825 | 0.8088 | 0.2080 | 0.0618 |
0.4472 | 33.0 | 825 | 0.4159 | 0.825 | 0.2625 | 1.5069 | 0.825 | 0.8135 | 0.2082 | 0.0604 |
0.4472 | 34.0 | 850 | 0.4168 | 0.825 | 0.2638 | 1.5770 | 0.825 | 0.8137 | 0.1888 | 0.0588 |
0.4472 | 35.0 | 875 | 0.4181 | 0.82 | 0.2640 | 1.5404 | 0.82 | 0.8043 | 0.2145 | 0.0582 |
0.4472 | 36.0 | 900 | 0.4154 | 0.83 | 0.2618 | 1.5719 | 0.83 | 0.8165 | 0.1965 | 0.0586 |
0.4472 | 37.0 | 925 | 0.4160 | 0.825 | 0.2632 | 1.5840 | 0.825 | 0.8137 | 0.2003 | 0.0604 |
0.4472 | 38.0 | 950 | 0.4133 | 0.83 | 0.2616 | 1.5711 | 0.83 | 0.8163 | 0.2040 | 0.0596 |
0.4472 | 39.0 | 975 | 0.4167 | 0.825 | 0.2635 | 1.5210 | 0.825 | 0.8138 | 0.1930 | 0.0590 |
0.0652 | 40.0 | 1000 | 0.4162 | 0.83 | 0.2630 | 1.6312 | 0.83 | 0.8163 | 0.1973 | 0.0593 |
0.0652 | 41.0 | 1025 | 0.4144 | 0.83 | 0.2626 | 1.5787 | 0.83 | 0.8163 | 0.2068 | 0.0603 |
0.0652 | 42.0 | 1050 | 0.4150 | 0.83 | 0.2631 | 1.5789 | 0.83 | 0.8163 | 0.1970 | 0.0588 |
0.0652 | 43.0 | 1075 | 0.4158 | 0.825 | 0.2635 | 1.5833 | 0.825 | 0.8138 | 0.1927 | 0.0597 |
0.0652 | 44.0 | 1100 | 0.4132 | 0.83 | 0.2622 | 1.5130 | 0.83 | 0.8163 | 0.2030 | 0.0593 |
0.0652 | 45.0 | 1125 | 0.4146 | 0.83 | 0.2630 | 1.6312 | 0.83 | 0.8165 | 0.2010 | 0.0587 |
0.0652 | 46.0 | 1150 | 0.4138 | 0.825 | 0.2624 | 1.5301 | 0.825 | 0.8135 | 0.2065 | 0.0587 |
0.0652 | 47.0 | 1175 | 0.4142 | 0.83 | 0.2627 | 1.6292 | 0.83 | 0.8163 | 0.1984 | 0.0591 |
0.0652 | 48.0 | 1200 | 0.4146 | 0.825 | 0.2629 | 1.5735 | 0.825 | 0.8137 | 0.1998 | 0.0589 |
0.0652 | 49.0 | 1225 | 0.4143 | 0.83 | 0.2630 | 1.5276 | 0.83 | 0.8163 | 0.2116 | 0.0599 |
0.0652 | 50.0 | 1250 | 0.4140 | 0.83 | 0.2628 | 1.5705 | 0.83 | 0.8163 | 0.1966 | 0.0590 |
0.0652 | 51.0 | 1275 | 0.4152 | 0.825 | 0.2637 | 1.5747 | 0.825 | 0.8138 | 0.1835 | 0.0593 |
0.0652 | 52.0 | 1300 | 0.4145 | 0.825 | 0.2629 | 1.5796 | 0.825 | 0.8137 | 0.1926 | 0.0593 |
0.0652 | 53.0 | 1325 | 0.4147 | 0.825 | 0.2631 | 1.6323 | 0.825 | 0.8138 | 0.1838 | 0.0588 |
0.0652 | 54.0 | 1350 | 0.4141 | 0.83 | 0.2628 | 1.5763 | 0.83 | 0.8163 | 0.2035 | 0.0592 |
0.0652 | 55.0 | 1375 | 0.4137 | 0.83 | 0.2630 | 1.5751 | 0.83 | 0.8163 | 0.2042 | 0.0590 |
0.0652 | 56.0 | 1400 | 0.4145 | 0.83 | 0.2632 | 1.6307 | 0.83 | 0.8163 | 0.1981 | 0.0588 |
0.0652 | 57.0 | 1425 | 0.4149 | 0.825 | 0.2634 | 1.5225 | 0.825 | 0.8137 | 0.2008 | 0.0589 |
0.0652 | 58.0 | 1450 | 0.4146 | 0.83 | 0.2634 | 1.5725 | 0.83 | 0.8163 | 0.2121 | 0.0589 |
0.0652 | 59.0 | 1475 | 0.4142 | 0.83 | 0.2632 | 1.5214 | 0.83 | 0.8163 | 0.2028 | 0.0590 |
0.0614 | 60.0 | 1500 | 0.4145 | 0.83 | 0.2634 | 1.5237 | 0.83 | 0.8163 | 0.1981 | 0.0585 |
0.0614 | 61.0 | 1525 | 0.4142 | 0.83 | 0.2630 | 1.5710 | 0.83 | 0.8163 | 0.2070 | 0.0591 |
0.0614 | 62.0 | 1550 | 0.4139 | 0.825 | 0.2631 | 1.5733 | 0.825 | 0.8135 | 0.1986 | 0.0594 |
0.0614 | 63.0 | 1575 | 0.4139 | 0.825 | 0.2630 | 1.5813 | 0.825 | 0.8136 | 0.1984 | 0.0593 |
0.0614 | 64.0 | 1600 | 0.4138 | 0.83 | 0.2629 | 1.5729 | 0.83 | 0.8163 | 0.2035 | 0.0590 |
0.0614 | 65.0 | 1625 | 0.4139 | 0.825 | 0.2629 | 1.5715 | 0.825 | 0.8136 | 0.2026 | 0.0593 |
0.0614 | 66.0 | 1650 | 0.4136 | 0.825 | 0.2629 | 1.5768 | 0.825 | 0.8135 | 0.1988 | 0.0592 |
0.0614 | 67.0 | 1675 | 0.4139 | 0.825 | 0.2629 | 1.5709 | 0.825 | 0.8135 | 0.1987 | 0.0593 |
0.0614 | 68.0 | 1700 | 0.4143 | 0.825 | 0.2633 | 1.5744 | 0.825 | 0.8138 | 0.1896 | 0.0595 |
0.0614 | 69.0 | 1725 | 0.4142 | 0.825 | 0.2632 | 1.5752 | 0.825 | 0.8138 | 0.1896 | 0.0593 |
0.0614 | 70.0 | 1750 | 0.4142 | 0.825 | 0.2632 | 1.5769 | 0.825 | 0.8138 | 0.1879 | 0.0594 |
0.0614 | 71.0 | 1775 | 0.4138 | 0.83 | 0.2630 | 1.5734 | 0.83 | 0.8163 | 0.2073 | 0.0588 |
0.0614 | 72.0 | 1800 | 0.4140 | 0.825 | 0.2631 | 1.5734 | 0.825 | 0.8138 | 0.1977 | 0.0593 |
0.0614 | 73.0 | 1825 | 0.4135 | 0.83 | 0.2629 | 1.5711 | 0.83 | 0.8163 | 0.2035 | 0.0589 |
0.0614 | 74.0 | 1850 | 0.4140 | 0.83 | 0.2632 | 1.5717 | 0.83 | 0.8163 | 0.2038 | 0.0590 |
0.0614 | 75.0 | 1875 | 0.4141 | 0.825 | 0.2633 | 1.5205 | 0.825 | 0.8138 | 0.1838 | 0.0593 |
0.0614 | 76.0 | 1900 | 0.4138 | 0.825 | 0.2631 | 1.5218 | 0.825 | 0.8137 | 0.1838 | 0.0595 |
0.0614 | 77.0 | 1925 | 0.4134 | 0.825 | 0.2628 | 1.5710 | 0.825 | 0.8135 | 0.1937 | 0.0591 |
0.0614 | 78.0 | 1950 | 0.4135 | 0.83 | 0.2629 | 1.5688 | 0.83 | 0.8163 | 0.2067 | 0.0588 |
0.0614 | 79.0 | 1975 | 0.4138 | 0.825 | 0.2631 | 1.5143 | 0.825 | 0.8137 | 0.1942 | 0.0592 |
0.0613 | 80.0 | 2000 | 0.4134 | 0.825 | 0.2628 | 1.5152 | 0.825 | 0.8135 | 0.1939 | 0.0591 |
0.0613 | 81.0 | 2025 | 0.4139 | 0.825 | 0.2632 | 1.5144 | 0.825 | 0.8136 | 0.1903 | 0.0593 |
0.0613 | 82.0 | 2050 | 0.4139 | 0.83 | 0.2632 | 1.5242 | 0.83 | 0.8163 | 0.1894 | 0.0589 |
0.0613 | 83.0 | 2075 | 0.4138 | 0.825 | 0.2631 | 1.5159 | 0.825 | 0.8136 | 0.2014 | 0.0594 |
0.0613 | 84.0 | 2100 | 0.4137 | 0.825 | 0.2631 | 1.5707 | 0.825 | 0.8136 | 0.1954 | 0.0592 |
0.0613 | 85.0 | 2125 | 0.4136 | 0.825 | 0.2630 | 1.5252 | 0.825 | 0.8136 | 0.1878 | 0.0592 |
0.0613 | 86.0 | 2150 | 0.4138 | 0.83 | 0.2630 | 1.5186 | 0.83 | 0.8172 | 0.2024 | 0.0588 |
0.0613 | 87.0 | 2175 | 0.4139 | 0.825 | 0.2632 | 1.5201 | 0.825 | 0.8138 | 0.1927 | 0.0592 |
0.0613 | 88.0 | 2200 | 0.4138 | 0.83 | 0.2631 | 1.5285 | 0.83 | 0.8172 | 0.1897 | 0.0591 |
0.0613 | 89.0 | 2225 | 0.4137 | 0.825 | 0.2631 | 1.5185 | 0.825 | 0.8136 | 0.1956 | 0.0593 |
0.0613 | 90.0 | 2250 | 0.4137 | 0.83 | 0.2631 | 1.5212 | 0.83 | 0.8172 | 0.2007 | 0.0591 |
0.0613 | 91.0 | 2275 | 0.4138 | 0.825 | 0.2631 | 1.5185 | 0.825 | 0.8138 | 0.1915 | 0.0593 |
0.0613 | 92.0 | 2300 | 0.4136 | 0.83 | 0.2630 | 1.5174 | 0.83 | 0.8172 | 0.2067 | 0.0590 |
0.0613 | 93.0 | 2325 | 0.4137 | 0.83 | 0.2631 | 1.5204 | 0.83 | 0.8172 | 0.1939 | 0.0591 |
0.0613 | 94.0 | 2350 | 0.4137 | 0.83 | 0.2631 | 1.5255 | 0.83 | 0.8172 | 0.2007 | 0.0592 |
0.0613 | 95.0 | 2375 | 0.4137 | 0.83 | 0.2631 | 1.5161 | 0.83 | 0.8172 | 0.1966 | 0.0591 |
0.0613 | 96.0 | 2400 | 0.4136 | 0.83 | 0.2630 | 1.5180 | 0.83 | 0.8172 | 0.2007 | 0.0590 |
0.0613 | 97.0 | 2425 | 0.4137 | 0.83 | 0.2631 | 1.5176 | 0.83 | 0.8172 | 0.1966 | 0.0591 |
0.0613 | 98.0 | 2450 | 0.4137 | 0.83 | 0.2631 | 1.5194 | 0.83 | 0.8172 | 0.1966 | 0.0590 |
0.0613 | 99.0 | 2475 | 0.4137 | 0.83 | 0.2631 | 1.5195 | 0.83 | 0.8172 | 0.2005 | 0.0591 |
0.0613 | 100.0 | 2500 | 0.4137 | 0.83 | 0.2631 | 1.5189 | 0.83 | 0.8172 | 0.2007 | 0.0591 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.13.1
- Datasets 2.13.1
- Tokenizers 0.13.3