vimassaru/teeth-seg-upernet-convnext-tiny

<!-- 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. -->

teeth-seg-upernet-convnext-tiny

This model is a fine-tuned version of openmmlab/upernet-convnext-tiny on the None dataset. It achieves the following results on the evaluation set:

• —Loss: 0.3710
• —Mean Iou: 0.7871
• —Mean Accuracy: 0.8577
• —Overall Accuracy: 0.8628
• —Accuracy Background: nan
• —Accuracy 11: 0.9002
• —Accuracy 12: 0.8915
• —Accuracy 13: 0.8824
• —Accuracy 14: 0.8296
• —Accuracy 15: 0.8450
• —Accuracy 16: 0.9027
• —Accuracy 17: 0.8757
• —Accuracy 18: 0.8080
• —Accuracy 21: 0.9085
• —Accuracy 22: 0.8828
• —Accuracy 23: 0.9049
• —Accuracy 24: 0.8117
• —Accuracy 25: 0.8581
• —Accuracy 26: 0.8913
• —Accuracy 27: 0.8817
• —Accuracy 28: 0.8016
• —Accuracy 31: 0.8020
• —Accuracy 32: 0.8415
• —Accuracy 33: 0.9144
• —Accuracy 34: 0.8550
• —Accuracy 35: 0.8463
• —Accuracy 36: 0.7861
• —Accuracy 37: 0.8171
• —Accuracy 38: 0.7589
• —Accuracy 41: 0.7945
• —Accuracy 42: 0.8365
• —Accuracy 43: 0.9052
• —Accuracy 44: 0.8953
• —Accuracy 45: 0.8821
• —Accuracy 46: 0.9060
• —Accuracy 47: 0.8907
• —Accuracy 48: 0.8379
• —Iou Background: 0.0
• —Iou 11: 0.8834
• —Iou 12: 0.8894
• —Iou 13: 0.8289
• —Iou 14: 0.7931
• —Iou 15: 0.8201
• —Iou 16: 0.8654
• —Iou 17: 0.8121
• —Iou 18: 0.7384
• —Iou 21: 0.8850
• —Iou 22: 0.8611
• —Iou 23: 0.8692
• —Iou 24: 0.7985
• —Iou 25: 0.8421
• —Iou 26: 0.8666
• —Iou 27: 0.7984
• —Iou 28: 0.7187
• —Iou 31: 0.7352
• —Iou 32: 0.8019
• —Iou 33: 0.8585
• —Iou 34: 0.8109
• —Iou 35: 0.7914
• —Iou 36: 0.7229
• —Iou 37: 0.7222
• —Iou 38: 0.6444
• —Iou 41: 0.7315
• —Iou 42: 0.7954
• —Iou 43: 0.8643
• —Iou 44: 0.8586
• —Iou 45: 0.8474
• —Iou 46: 0.8721
• —Iou 47: 0.8579
• —Iou 48: 0.7878

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: 0.0006
• —trainbatchsize: 8
• —evalbatchsize: 8
• —seed: 42
• —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• —lrschedulertype: linear
• —num_epochs: 200

Training results

Framework versions

• —Transformers 4.30.2
• —Pytorch 2.0.1+cu118
• —Datasets 2.13.1
• —Tokenizers 0.13.3