--- license: apache-2.0 base_model: google/vit-base-patch16-224-in21k tags: - image-classification - vision - generated_from_trainer datasets: - imagefolder metrics: - accuracy model-index: - name: vit-letter-identification-v2 results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.8627450980392157 --- # vit-letter-identification-v2 This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset. It achieves the following results on the evaluation set: - Loss: 1.1135 - Accuracy: 0.8627 ## 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: 100 - eval_batch_size: 102 - seed: 1337 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 120.0 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | No log | 1.0 | 6 | 3.2331 | 0.0882 | | 3.2363 | 2.0 | 12 | 3.2025 | 0.1373 | | 3.2363 | 3.0 | 18 | 3.1761 | 0.1863 | | 3.1622 | 4.0 | 24 | 3.1238 | 0.2255 | | 3.0918 | 5.0 | 30 | 3.0789 | 0.3137 | | 3.0918 | 6.0 | 36 | 3.0280 | 0.3235 | | 3.0081 | 7.0 | 42 | 2.9878 | 0.3431 | | 3.0081 | 8.0 | 48 | 2.9316 | 0.3824 | | 2.9118 | 9.0 | 54 | 2.8864 | 0.4314 | | 2.8231 | 10.0 | 60 | 2.8314 | 0.4510 | | 2.8231 | 11.0 | 66 | 2.7817 | 0.5196 | | 2.7149 | 12.0 | 72 | 2.7278 | 0.5196 | | 2.7149 | 13.0 | 78 | 2.6796 | 0.5588 | | 2.6202 | 14.0 | 84 | 2.6203 | 0.5882 | | 2.5243 | 15.0 | 90 | 2.5674 | 0.5882 | | 2.5243 | 16.0 | 96 | 2.5170 | 0.6078 | | 2.4279 | 17.0 | 102 | 2.4672 | 0.6176 | | 2.4279 | 18.0 | 108 | 2.4285 | 0.5980 | | 2.3404 | 19.0 | 114 | 2.3784 | 0.6569 | | 2.2633 | 20.0 | 120 | 2.3348 | 0.6471 | | 2.2633 | 21.0 | 126 | 2.2872 | 0.6667 | | 2.1838 | 22.0 | 132 | 2.2539 | 0.6569 | | 2.1838 | 23.0 | 138 | 2.2232 | 0.6765 | | 2.1022 | 24.0 | 144 | 2.1867 | 0.6471 | | 2.0364 | 25.0 | 150 | 2.1489 | 0.6863 | | 2.0364 | 26.0 | 156 | 2.1099 | 0.7255 | | 1.96 | 27.0 | 162 | 2.0767 | 0.7157 | | 1.96 | 28.0 | 168 | 2.0417 | 0.7157 | | 1.9235 | 29.0 | 174 | 2.0162 | 0.7353 | | 1.8484 | 30.0 | 180 | 1.9787 | 0.7451 | | 1.8484 | 31.0 | 186 | 1.9548 | 0.7451 | | 1.7971 | 32.0 | 192 | 1.9329 | 0.7549 | | 1.7971 | 33.0 | 198 | 1.9052 | 0.7647 | | 1.7409 | 34.0 | 204 | 1.8827 | 0.7549 | | 1.7006 | 35.0 | 210 | 1.8589 | 0.7745 | | 1.7006 | 36.0 | 216 | 1.8294 | 0.7843 | | 1.6426 | 37.0 | 222 | 1.8098 | 0.7843 | | 1.6426 | 38.0 | 228 | 1.7809 | 0.7647 | | 1.6102 | 39.0 | 234 | 1.7643 | 0.7843 | | 1.5704 | 40.0 | 240 | 1.7399 | 0.8039 | | 1.5704 | 41.0 | 246 | 1.7193 | 0.8137 | | 1.5264 | 42.0 | 252 | 1.6980 | 0.8333 | | 1.5264 | 43.0 | 258 | 1.6840 | 0.8039 | | 1.4821 | 44.0 | 264 | 1.6644 | 0.8235 | | 1.4506 | 45.0 | 270 | 1.6467 | 0.8235 | | 1.4506 | 46.0 | 276 | 1.6333 | 0.8235 | | 1.4358 | 47.0 | 282 | 1.6095 | 0.8235 | | 1.4358 | 48.0 | 288 | 1.5906 | 0.8235 | | 1.3695 | 49.0 | 294 | 1.5720 | 0.8431 | | 1.367 | 50.0 | 300 | 1.5610 | 0.8333 | | 1.367 | 51.0 | 306 | 1.5440 | 0.8529 | | 1.3299 | 52.0 | 312 | 1.5359 | 0.8333 | | 1.3299 | 53.0 | 318 | 1.5129 | 0.8333 | | 1.2765 | 54.0 | 324 | 1.5057 | 0.8235 | | 1.2785 | 55.0 | 330 | 1.4867 | 0.8235 | | 1.2785 | 56.0 | 336 | 1.4751 | 0.8333 | | 1.2355 | 57.0 | 342 | 1.4553 | 0.8235 | | 1.2355 | 58.0 | 348 | 1.4491 | 0.8235 | | 1.2418 | 59.0 | 354 | 1.4289 | 0.8431 | | 1.2058 | 60.0 | 360 | 1.4185 | 0.8235 | | 1.2058 | 61.0 | 366 | 1.4104 | 0.8333 | | 1.164 | 62.0 | 372 | 1.3968 | 0.8333 | | 1.164 | 63.0 | 378 | 1.3846 | 0.8431 | | 1.1529 | 64.0 | 384 | 1.3697 | 0.8431 | | 1.1408 | 65.0 | 390 | 1.3633 | 0.8431 | | 1.1408 | 66.0 | 396 | 1.3505 | 0.8431 | | 1.1102 | 67.0 | 402 | 1.3371 | 0.8529 | | 1.1102 | 68.0 | 408 | 1.3282 | 0.8529 | | 1.0906 | 69.0 | 414 | 1.3240 | 0.8431 | | 1.0759 | 70.0 | 420 | 1.3163 | 0.8431 | | 1.0759 | 71.0 | 426 | 1.3044 | 0.8529 | | 1.0651 | 72.0 | 432 | 1.2924 | 0.8431 | | 1.0651 | 73.0 | 438 | 1.2867 | 0.8529 | | 1.0501 | 74.0 | 444 | 1.2749 | 0.8529 | | 1.0238 | 75.0 | 450 | 1.2688 | 0.8431 | | 1.0238 | 76.0 | 456 | 1.2568 | 0.8529 | | 1.0046 | 77.0 | 462 | 1.2502 | 0.8529 | | 1.0046 | 78.0 | 468 | 1.2460 | 0.8529 | | 0.9946 | 79.0 | 474 | 1.2455 | 0.8431 | | 0.9998 | 80.0 | 480 | 1.2343 | 0.8529 | | 0.9998 | 81.0 | 486 | 1.2286 | 0.8529 | | 0.9709 | 82.0 | 492 | 1.2195 | 0.8431 | | 0.9709 | 83.0 | 498 | 1.2126 | 0.8529 | | 0.963 | 84.0 | 504 | 1.2102 | 0.8431 | | 0.9499 | 85.0 | 510 | 1.2024 | 0.8431 | | 0.9499 | 86.0 | 516 | 1.1980 | 0.8529 | | 0.937 | 87.0 | 522 | 1.1912 | 0.8529 | | 0.937 | 88.0 | 528 | 1.1883 | 0.8431 | | 0.9389 | 89.0 | 534 | 1.1845 | 0.8529 | | 0.9181 | 90.0 | 540 | 1.1811 | 0.8529 | | 0.9181 | 91.0 | 546 | 1.1777 | 0.8431 | | 0.9219 | 92.0 | 552 | 1.1743 | 0.8627 | | 0.9219 | 93.0 | 558 | 1.1675 | 0.8627 | | 0.9067 | 94.0 | 564 | 1.1598 | 0.8627 | | 0.9009 | 95.0 | 570 | 1.1601 | 0.8627 | | 0.9009 | 96.0 | 576 | 1.1564 | 0.8529 | | 0.8914 | 97.0 | 582 | 1.1505 | 0.8529 | | 0.8914 | 98.0 | 588 | 1.1487 | 0.8529 | | 0.8739 | 99.0 | 594 | 1.1480 | 0.8627 | | 0.8742 | 100.0 | 600 | 1.1413 | 0.8529 | | 0.8742 | 101.0 | 606 | 1.1368 | 0.8627 | | 0.8679 | 102.0 | 612 | 1.1361 | 0.8627 | | 0.8679 | 103.0 | 618 | 1.1317 | 0.8627 | | 0.8516 | 104.0 | 624 | 1.1296 | 0.8529 | | 0.876 | 105.0 | 630 | 1.1288 | 0.8627 | | 0.876 | 106.0 | 636 | 1.1264 | 0.8627 | | 0.8591 | 107.0 | 642 | 1.1238 | 0.8627 | | 0.8591 | 108.0 | 648 | 1.1227 | 0.8627 | | 0.8586 | 109.0 | 654 | 1.1208 | 0.8627 | | 0.8415 | 110.0 | 660 | 1.1194 | 0.8627 | | 0.8415 | 111.0 | 666 | 1.1185 | 0.8627 | | 0.8465 | 112.0 | 672 | 1.1178 | 0.8529 | | 0.8465 | 113.0 | 678 | 1.1184 | 0.8529 | | 0.8503 | 114.0 | 684 | 1.1183 | 0.8431 | | 0.8332 | 115.0 | 690 | 1.1174 | 0.8431 | | 0.8332 | 116.0 | 696 | 1.1165 | 0.8431 | | 0.8476 | 117.0 | 702 | 1.1153 | 0.8529 | | 0.8476 | 118.0 | 708 | 1.1142 | 0.8529 | | 0.8382 | 119.0 | 714 | 1.1137 | 0.8627 | | 0.8527 | 120.0 | 720 | 1.1135 | 0.8627 | ### Framework versions - Transformers 4.37.0.dev0 - Pytorch 2.1.0+cu121 - Datasets 2.15.0 - Tokenizers 0.15.0