Edit model card

swin-tiny-patch4-window7-224-ve-U11-b-40

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6121
  • Accuracy: 0.8261

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: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 40

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 0.92 6 1.5799 0.4783
2.1773 2.0 13 1.5648 0.3478
2.1773 2.92 19 1.5182 0.3261
2.1773 4.0 26 1.4232 0.3261
1.8993 4.92 32 1.3505 0.3913
1.8993 6.0 39 1.2747 0.3696
1.5045 6.92 45 1.2452 0.3696
1.2431 8.0 52 1.1982 0.2826
1.2431 8.92 58 1.2112 0.3043
1.1225 10.0 65 1.0160 0.5
0.9942 10.92 71 1.0138 0.4783
0.9942 12.0 78 0.9094 0.5652
0.9212 12.92 84 0.8860 0.5217
0.816 14.0 91 0.7693 0.6739
0.816 14.92 97 0.8290 0.6304
0.741 16.0 104 0.7810 0.6739
0.631 16.92 110 0.6342 0.7826
0.631 18.0 117 0.7677 0.6957
0.6402 18.92 123 0.6283 0.7391
0.5477 20.0 130 0.6687 0.7174
0.5477 20.92 136 0.6369 0.7826
0.5023 22.0 143 0.6334 0.7609
0.5023 22.92 149 0.6355 0.8043
0.4802 24.0 156 0.5976 0.8043
0.4336 24.92 162 0.6112 0.7609
0.4336 26.0 169 0.6148 0.8043
0.4203 26.92 175 0.6380 0.7391
0.429 28.0 182 0.6032 0.8043
0.429 28.92 188 0.6348 0.7391
0.4013 30.0 195 0.6121 0.8261
0.3747 30.92 201 0.6521 0.7391
0.3747 32.0 208 0.6424 0.7609
0.3668 32.92 214 0.6149 0.8261
0.3287 34.0 221 0.6426 0.7826
0.3287 34.92 227 0.6379 0.8043
0.372 36.0 234 0.6435 0.8043
0.3236 36.92 240 0.6450 0.8043

Framework versions

  • Transformers 4.36.2
  • Pytorch 2.1.2+cu118
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
2
Safetensors
Model size
27.6M params
Tensor type
I64
·
F32
·
Inference API
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 Augusto777/swin-tiny-patch4-window7-224-ve-U11-b-40

Finetuned
this model

Evaluation results