Hugging Face's logo

AmirGenAI
/
segformer-b0-finetuned-wrinkle

Image Segmentation
Transformers
TensorBoard
Safetensors
segformer
vision
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community

segformer-b0-finetuned-wrinkle

This model is a fine-tuned version of nvidia/mit-b0 on the face-wrinkles dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0189
  • Mean Iou: 0.2163
  • Mean Accuracy: 0.4327
  • Overall Accuracy: 0.4327
  • Accuracy Unlabeled: nan
  • Accuracy Wrinkle: 0.4327
  • Iou Unlabeled: 0.0
  • Iou Wrinkle: 0.4327

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: 6e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • 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
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Mean Iou Mean Accuracy Overall Accuracy Accuracy Unlabeled Accuracy Wrinkle Iou Unlabeled Iou Wrinkle
0.0122 0.1786 20 0.0186 0.1899 0.3798 0.3798 nan 0.3798 0.0 0.3798
0.0114 0.3571 40 0.0188 0.2007 0.4014 0.4014 nan 0.4014 0.0 0.4014
0.0104 0.5357 60 0.0189 0.2127 0.4254 0.4254 nan 0.4254 0.0 0.4254
0.0116 0.7143 80 0.0187 0.2215 0.4430 0.4430 nan 0.4430 0.0 0.4430
0.0104 0.8929 100 0.0189 0.1815 0.3630 0.3630 nan 0.3630 0.0 0.3630
0.0151 1.0714 120 0.0187 0.1949 0.3898 0.3898 nan 0.3898 0.0 0.3898
0.0155 1.25 140 0.0187 0.2073 0.4147 0.4147 nan 0.4147 0.0 0.4147
0.0077 1.4286 160 0.0192 0.2406 0.4812 0.4812 nan 0.4812 0.0 0.4812
0.0117 1.6071 180 0.0191 0.2391 0.4782 0.4782 nan 0.4782 0.0 0.4782
0.0063 1.7857 200 0.0188 0.1787 0.3573 0.3573 nan 0.3573 0.0 0.3573
0.01 1.9643 220 0.0185 0.2195 0.4389 0.4389 nan 0.4389 0.0 0.4389
0.0109 2.1429 240 0.0191 0.1699 0.3398 0.3398 nan 0.3398 0.0 0.3398
0.0104 2.3214 260 0.0191 0.2167 0.4335 0.4335 nan 0.4335 0.0 0.4335
0.0145 2.5 280 0.0198 0.2604 0.5208 0.5208 nan 0.5208 0.0 0.5208
0.0093 2.6786 300 0.0185 0.1963 0.3927 0.3927 nan 0.3927 0.0 0.3927
0.0106 2.8571 320 0.0185 0.2080 0.4159 0.4159 nan 0.4159 0.0 0.4159
0.007 3.0357 340 0.0190 0.1894 0.3787 0.3787 nan 0.3787 0.0 0.3787
0.01 3.2143 360 0.0189 0.2194 0.4389 0.4389 nan 0.4389 0.0 0.4389
0.0118 3.3929 380 0.0186 0.2312 0.4625 0.4625 nan 0.4625 0.0 0.4625
0.008 3.5714 400 0.0189 0.1746 0.3492 0.3492 nan 0.3492 0.0 0.3492
0.0101 3.75 420 0.0185 0.1822 0.3644 0.3644 nan 0.3644 0.0 0.3644
0.0093 3.9286 440 0.0187 0.2126 0.4252 0.4252 nan 0.4252 0.0 0.4252
0.008 4.1071 460 0.0186 0.2058 0.4116 0.4116 nan 0.4116 0.0 0.4116
0.0134 4.2857 480 0.0187 0.2335 0.4669 0.4669 nan 0.4669 0.0 0.4669
0.0119 4.4643 500 0.0191 0.1850 0.3700 0.3700 nan 0.3700 0.0 0.3700
0.0064 4.6429 520 0.0187 0.1892 0.3785 0.3785 nan 0.3785 0.0 0.3785
0.0087 4.8214 540 0.0190 0.2253 0.4506 0.4506 nan 0.4506 0.0 0.4506
0.0122 5.0 560 0.0196 0.2598 0.5196 0.5196 nan 0.5196 0.0 0.5196
0.0071 5.1786 580 0.0188 0.2224 0.4448 0.4448 nan 0.4448 0.0 0.4448
0.0125 5.3571 600 0.0188 0.2051 0.4103 0.4103 nan 0.4103 0.0 0.4103
0.0093 5.5357 620 0.0192 0.2410 0.4821 0.4821 nan 0.4821 0.0 0.4821
0.0082 5.7143 640 0.0191 0.2291 0.4582 0.4582 nan 0.4582 0.0 0.4582
0.0089 5.8929 660 0.0187 0.1993 0.3985 0.3985 nan 0.3985 0.0 0.3985
0.0104 6.0714 680 0.0191 0.2049 0.4098 0.4098 nan 0.4098 0.0 0.4098
0.0111 6.25 700 0.0187 0.2216 0.4431 0.4431 nan 0.4431 0.0 0.4431
0.0113 6.4286 720 0.0196 0.2525 0.5050 0.5050 nan 0.5050 0.0 0.5050
0.0099 6.6071 740 0.0189 0.2219 0.4439 0.4439 nan 0.4439 0.0 0.4439
0.0062 6.7857 760 0.0187 0.2349 0.4699 0.4699 nan 0.4699 0.0 0.4699
0.0132 6.9643 780 0.0188 0.2108 0.4217 0.4217 nan 0.4217 0.0 0.4217
0.0132 7.1429 800 0.0190 0.2097 0.4194 0.4194 nan 0.4194 0.0 0.4194
0.0141 7.3214 820 0.0187 0.2125 0.4251 0.4251 nan 0.4251 0.0 0.4251
0.0121 7.5 840 0.0189 0.2176 0.4351 0.4351 nan 0.4351 0.0 0.4351
0.0099 7.6786 860 0.0187 0.2002 0.4004 0.4004 nan 0.4004 0.0 0.4004
0.0168 7.8571 880 0.0188 0.2159 0.4319 0.4319 nan 0.4319 0.0 0.4319
0.0064 8.0357 900 0.0188 0.2194 0.4387 0.4387 nan 0.4387 0.0 0.4387
0.0121 8.2143 920 0.0191 0.2309 0.4618 0.4618 nan 0.4618 0.0 0.4618
0.0133 8.3929 940 0.0189 0.2101 0.4202 0.4202 nan 0.4202 0.0 0.4202
0.0105 8.5714 960 0.0190 0.2287 0.4573 0.4573 nan 0.4573 0.0 0.4573
0.0092 8.75 980 0.0188 0.2178 0.4356 0.4356 nan 0.4356 0.0 0.4356
0.0124 8.9286 1000 0.0191 0.2277 0.4553 0.4553 nan 0.4553 0.0 0.4553
0.0108 9.1071 1020 0.0189 0.2017 0.4033 0.4033 nan 0.4033 0.0 0.4033
0.0098 9.2857 1040 0.0190 0.2271 0.4542 0.4542 nan 0.4542 0.0 0.4542
0.0087 9.4643 1060 0.0189 0.2168 0.4335 0.4335 nan 0.4335 0.0 0.4335
0.008 9.6429 1080 0.0189 0.2219 0.4438 0.4438 nan 0.4438 0.0 0.4438
0.0071 9.8214 1100 0.0189 0.2204 0.4407 0.4407 nan 0.4407 0.0 0.4407
0.0072 10.0 1120 0.0189 0.2163 0.4327 0.4327 nan 0.4327 0.0 0.4327

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.5.1+cu121
  • Datasets 3.2.0
  • Tokenizers 0.20.3
Downloads last month
65
Safetensors
Model size
3.72M params
Tensor type
F32
·
Inference Examples
Image Segmentation
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for AmirGenAI/segformer-b0-finetuned-wrinkle

Base model

nvidia/mit-b0
Finetuned
(335)
this model

Evaluation results

Metadata error: specify a dataset to view leaderboard