--- library_name: transformers license: apache-2.0 base_model: google/vit-base-patch16-224-in21k tags: - generated_from_trainer datasets: - imagefolder metrics: - accuracy - precision - recall - f1 model-index: - name: vit-base-kidney-stone-5-Michel_Daudon_-w256_1k_v1-_MIX results: - task: name: Image Classification type: image-classification dataset: name: imagefolder type: imagefolder config: default split: test args: default metrics: - name: Accuracy type: accuracy value: 0.88875 - name: Precision type: precision value: 0.8974543575746378 - name: Recall type: recall value: 0.88875 - name: F1 type: f1 value: 0.8871125111810537 --- # vit-base-kidney-stone-5-Michel_Daudon_-w256_1k_v1-_MIX 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: 0.3946 - Accuracy: 0.8888 - Precision: 0.8975 - Recall: 0.8888 - F1: 0.8871 ## 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: 0.0002 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments - lr_scheduler_type: linear - num_epochs: 15 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |:-------------:|:-------:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| | 0.5771 | 0.1667 | 100 | 0.6379 | 0.7929 | 0.8436 | 0.7929 | 0.7925 | | 0.3294 | 0.3333 | 200 | 0.7346 | 0.7992 | 0.8342 | 0.7992 | 0.7915 | | 0.5113 | 0.5 | 300 | 0.5429 | 0.8638 | 0.8829 | 0.8638 | 0.8625 | | 0.1584 | 0.6667 | 400 | 0.6327 | 0.8304 | 0.8612 | 0.8304 | 0.8308 | | 0.2638 | 0.8333 | 500 | 1.0157 | 0.7575 | 0.7964 | 0.7575 | 0.7623 | | 0.2057 | 1.0 | 600 | 0.3946 | 0.8888 | 0.8975 | 0.8888 | 0.8871 | | 0.1699 | 1.1667 | 700 | 0.7519 | 0.7987 | 0.8373 | 0.7987 | 0.8004 | | 0.1526 | 1.3333 | 800 | 0.7253 | 0.8342 | 0.8727 | 0.8342 | 0.8372 | | 0.0361 | 1.5 | 900 | 1.0151 | 0.7829 | 0.8064 | 0.7829 | 0.7748 | | 0.0756 | 1.6667 | 1000 | 0.6614 | 0.8625 | 0.8860 | 0.8625 | 0.8647 | | 0.0267 | 1.8333 | 1100 | 0.9163 | 0.8154 | 0.8321 | 0.8154 | 0.8195 | | 0.1447 | 2.0 | 1200 | 0.7084 | 0.8271 | 0.8381 | 0.8271 | 0.8244 | | 0.0132 | 2.1667 | 1300 | 0.8919 | 0.8354 | 0.8758 | 0.8354 | 0.8378 | | 0.0254 | 2.3333 | 1400 | 0.7531 | 0.8488 | 0.8772 | 0.8488 | 0.8505 | | 0.0848 | 2.5 | 1500 | 0.6491 | 0.8733 | 0.8841 | 0.8733 | 0.8765 | | 0.0605 | 2.6667 | 1600 | 0.7045 | 0.855 | 0.8708 | 0.855 | 0.8515 | | 0.0085 | 2.8333 | 1700 | 1.1652 | 0.7992 | 0.8305 | 0.7992 | 0.7879 | | 0.1798 | 3.0 | 1800 | 0.9389 | 0.8075 | 0.8350 | 0.8075 | 0.8075 | | 0.0555 | 3.1667 | 1900 | 0.7451 | 0.8421 | 0.8593 | 0.8421 | 0.8452 | | 0.0245 | 3.3333 | 2000 | 0.4729 | 0.8888 | 0.8942 | 0.8888 | 0.8880 | | 0.0017 | 3.5 | 2100 | 0.7608 | 0.8629 | 0.8859 | 0.8629 | 0.8663 | | 0.0266 | 3.6667 | 2200 | 0.7795 | 0.8571 | 0.8668 | 0.8571 | 0.8578 | | 0.0072 | 3.8333 | 2300 | 0.6487 | 0.8596 | 0.8862 | 0.8596 | 0.8600 | | 0.0019 | 4.0 | 2400 | 0.6297 | 0.8712 | 0.8846 | 0.8712 | 0.8723 | | 0.001 | 4.1667 | 2500 | 0.8346 | 0.8679 | 0.8849 | 0.8679 | 0.8692 | | 0.0014 | 4.3333 | 2600 | 0.8441 | 0.8633 | 0.8869 | 0.8633 | 0.8671 | | 0.0068 | 4.5 | 2700 | 0.7032 | 0.8662 | 0.8769 | 0.8662 | 0.8649 | | 0.0014 | 4.6667 | 2800 | 0.7379 | 0.86 | 0.8795 | 0.86 | 0.8565 | | 0.0951 | 4.8333 | 2900 | 0.5960 | 0.8979 | 0.9086 | 0.8979 | 0.8984 | | 0.0439 | 5.0 | 3000 | 0.6975 | 0.8708 | 0.8902 | 0.8708 | 0.8699 | | 0.1022 | 5.1667 | 3100 | 1.0231 | 0.8363 | 0.8703 | 0.8363 | 0.8312 | | 0.0239 | 5.3333 | 3200 | 0.7746 | 0.8683 | 0.8767 | 0.8683 | 0.8690 | | 0.0087 | 5.5 | 3300 | 0.8246 | 0.8567 | 0.8700 | 0.8567 | 0.8561 | | 0.001 | 5.6667 | 3400 | 1.0921 | 0.8237 | 0.8484 | 0.8237 | 0.8208 | | 0.0056 | 5.8333 | 3500 | 0.7431 | 0.8533 | 0.8562 | 0.8533 | 0.8524 | | 0.0007 | 6.0 | 3600 | 0.8992 | 0.8213 | 0.8463 | 0.8213 | 0.8270 | | 0.0041 | 6.1667 | 3700 | 0.8531 | 0.8438 | 0.8757 | 0.8438 | 0.8454 | | 0.0138 | 6.3333 | 3800 | 0.6643 | 0.8821 | 0.8918 | 0.8821 | 0.8809 | | 0.0005 | 6.5 | 3900 | 0.6779 | 0.8862 | 0.8970 | 0.8862 | 0.8877 | | 0.0005 | 6.6667 | 4000 | 0.7109 | 0.8892 | 0.9030 | 0.8892 | 0.8903 | | 0.0005 | 6.8333 | 4100 | 0.7191 | 0.8908 | 0.9013 | 0.8908 | 0.8911 | | 0.0006 | 7.0 | 4200 | 0.8573 | 0.8675 | 0.8846 | 0.8675 | 0.8635 | | 0.064 | 7.1667 | 4300 | 0.9180 | 0.8608 | 0.8743 | 0.8608 | 0.8603 | | 0.0005 | 7.3333 | 4400 | 0.7651 | 0.8767 | 0.8885 | 0.8767 | 0.8763 | | 0.0007 | 7.5 | 4500 | 0.8158 | 0.8571 | 0.8703 | 0.8571 | 0.8569 | | 0.0004 | 7.6667 | 4600 | 0.8329 | 0.8504 | 0.8709 | 0.8504 | 0.8517 | | 0.0003 | 7.8333 | 4700 | 0.9078 | 0.8454 | 0.8605 | 0.8454 | 0.8446 | | 0.0003 | 8.0 | 4800 | 0.8859 | 0.8529 | 0.8684 | 0.8529 | 0.8538 | | 0.0003 | 8.1667 | 4900 | 0.9303 | 0.8479 | 0.8669 | 0.8479 | 0.8491 | | 0.0002 | 8.3333 | 5000 | 0.9324 | 0.8475 | 0.8676 | 0.8475 | 0.8483 | | 0.0002 | 8.5 | 5100 | 0.9206 | 0.8533 | 0.8733 | 0.8533 | 0.8544 | | 0.0002 | 8.6667 | 5200 | 0.8745 | 0.8621 | 0.8813 | 0.8621 | 0.8630 | | 0.0002 | 8.8333 | 5300 | 0.9208 | 0.8567 | 0.8764 | 0.8567 | 0.8575 | | 0.0002 | 9.0 | 5400 | 0.9221 | 0.8583 | 0.8776 | 0.8583 | 0.8592 | | 0.0002 | 9.1667 | 5500 | 0.9255 | 0.8588 | 0.8777 | 0.8588 | 0.8596 | | 0.0002 | 9.3333 | 5600 | 0.9285 | 0.8583 | 0.8772 | 0.8583 | 0.8592 | | 0.0001 | 9.5 | 5700 | 0.9288 | 0.8592 | 0.8780 | 0.8592 | 0.8601 | | 0.0001 | 9.6667 | 5800 | 0.9305 | 0.8596 | 0.8782 | 0.8596 | 0.8605 | | 0.0002 | 9.8333 | 5900 | 0.9323 | 0.8596 | 0.8782 | 0.8596 | 0.8605 | | 0.0001 | 10.0 | 6000 | 0.9335 | 0.8596 | 0.8782 | 0.8596 | 0.8606 | | 0.0001 | 10.1667 | 6100 | 0.9336 | 0.8608 | 0.8791 | 0.8608 | 0.8619 | | 0.0001 | 10.3333 | 6200 | 0.9360 | 0.8612 | 0.8795 | 0.8612 | 0.8623 | | 0.0001 | 10.5 | 6300 | 0.9374 | 0.8625 | 0.8803 | 0.8625 | 0.8635 | | 0.0001 | 10.6667 | 6400 | 0.9406 | 0.8629 | 0.8809 | 0.8629 | 0.8640 | | 0.0001 | 10.8333 | 6500 | 0.9420 | 0.8633 | 0.8810 | 0.8633 | 0.8643 | | 0.0001 | 11.0 | 6600 | 0.9443 | 0.8633 | 0.8810 | 0.8633 | 0.8643 | | 0.0001 | 11.1667 | 6700 | 0.9452 | 0.8633 | 0.8810 | 0.8633 | 0.8643 | | 0.0001 | 11.3333 | 6800 | 0.9476 | 0.8638 | 0.8813 | 0.8638 | 0.8647 | | 0.0001 | 11.5 | 6900 | 0.9495 | 0.8638 | 0.8813 | 0.8638 | 0.8647 | | 0.0001 | 11.6667 | 7000 | 0.9501 | 0.8642 | 0.8818 | 0.8642 | 0.8652 | | 0.0001 | 11.8333 | 7100 | 0.9528 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.0 | 7200 | 0.9547 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.1667 | 7300 | 0.9574 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.3333 | 7400 | 0.9586 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.5 | 7500 | 0.9594 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.6667 | 7600 | 0.9611 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 12.8333 | 7700 | 0.9627 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 13.0 | 7800 | 0.9639 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 13.1667 | 7900 | 0.9656 | 0.8646 | 0.8820 | 0.8646 | 0.8656 | | 0.0001 | 13.3333 | 8000 | 0.9662 | 0.8646 | 0.8820 | 0.8646 | 0.8655 | | 0.0001 | 13.5 | 8100 | 0.9675 | 0.8642 | 0.8815 | 0.8642 | 0.8651 | | 0.0001 | 13.6667 | 8200 | 0.9684 | 0.8642 | 0.8814 | 0.8642 | 0.8651 | | 0.0001 | 13.8333 | 8300 | 0.9695 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.0 | 8400 | 0.9706 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.1667 | 8500 | 0.9714 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.3333 | 8600 | 0.9724 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.5 | 8700 | 0.9727 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.6667 | 8800 | 0.9733 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 14.8333 | 8900 | 0.9734 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | | 0.0001 | 15.0 | 9000 | 0.9736 | 0.8646 | 0.8818 | 0.8646 | 0.8656 | ### Framework versions - Transformers 4.48.2 - Pytorch 2.6.0+cu126 - Datasets 3.1.0 - Tokenizers 0.21.0