metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: swin-tiny-patch4-window7-224-uploads-classifier-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: 0.984313725490196

swin-tiny-patch4-window7-224-uploads-classifier-v2

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: 0.0745
Accuracy: 0.9843

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.2482	1.0	18	0.4781	0.8824
0.3036	2.0	36	0.0936	0.9804
0.1687	3.0	54	0.0745	0.9843
0.1392	4.0	72	0.0980	0.9725
0.14	5.0	90	0.0778	0.9765
0.1186	6.0	108	0.0837	0.9725
0.1088	7.0	126	0.0645	0.9804
0.0789	8.0	144	0.0675	0.9765
0.0644	9.0	162	0.0940	0.9686
0.0582	10.0	180	0.0879	0.9725
0.0591	11.0	198	0.0935	0.9686
0.0538	12.0	216	0.0540	0.9804
0.0588	13.0	234	0.0725	0.9686
0.0538	14.0	252	0.0637	0.9765
0.0462	15.0	270	0.0694	0.9725
0.0352	16.0	288	0.0771	0.9686
0.0536	17.0	306	0.0629	0.9804
0.0403	18.0	324	0.0933	0.9686
0.0412	19.0	342	0.0848	0.9725
0.0305	20.0	360	0.0820	0.9725

Framework versions

Transformers 4.28.0
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.13.3