metadata

license: apache-2.0
base_model: microsoft/swin-tiny-patch4-window7-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
  - precision
  - recall
model-index:
  - name: swin-tiny-patch4-window7-224
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: validation
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8033333333333333
          - name: Precision
            type: precision
            value: 0.7970708748615725
          - name: Recall
            type: recall
            value: 0.8033333333333333

swin-tiny-patch4-window7-224

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: 0.4788
Accuracy: 0.8033
Precision: 0.7971
Recall: 0.8033
F1 Score: 0.7802

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: 64
eval_batch_size: 64
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 256
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1 Score
No log	1.0	4	0.5946	0.7333	0.5378	0.7333	0.6205
No log	2.0	8	0.6006	0.7333	0.5378	0.7333	0.6205
No log	3.0	12	0.5677	0.7333	0.5378	0.7333	0.6205
No log	4.0	16	0.5616	0.7333	0.5378	0.7333	0.6205
No log	5.0	20	0.5556	0.75	0.7193	0.75	0.7023
No log	6.0	24	0.5435	0.7667	0.7819	0.7667	0.7019
No log	7.0	28	0.5318	0.7792	0.7885	0.7792	0.7281
0.5745	8.0	32	0.5316	0.7542	0.7262	0.7542	0.7126
0.5745	9.0	36	0.5232	0.7667	0.7533	0.7667	0.7185
0.5745	10.0	40	0.5226	0.7708	0.7639	0.7708	0.7217
0.5745	11.0	44	0.5217	0.7708	0.7597	0.7708	0.7253
0.5745	12.0	48	0.5224	0.7625	0.7561	0.7625	0.7034
0.5745	13.0	52	0.5213	0.7708	0.7510	0.7708	0.7409
0.5745	14.0	56	0.5207	0.7667	0.7709	0.7667	0.7064
0.4741	15.0	60	0.5247	0.7583	0.7343	0.7583	0.7334
0.4741	16.0	64	0.5352	0.7708	0.7639	0.7708	0.7217
0.4741	17.0	68	0.5227	0.7708	0.7507	0.7708	0.7460
0.4741	18.0	72	0.5206	0.7583	0.7564	0.7583	0.6912
0.4741	19.0	76	0.5088	0.775	0.7627	0.775	0.7353
0.4741	20.0	80	0.5144	0.7667	0.7503	0.7667	0.7221
0.4741	21.0	84	0.5227	0.7875	0.7918	0.7875	0.7453
0.4741	22.0	88	0.5150	0.775	0.7564	0.775	0.7494
0.4233	23.0	92	0.5240	0.7667	0.7533	0.7667	0.7185
0.4233	24.0	96	0.5156	0.7792	0.7684	0.7792	0.7418
0.4233	25.0	100	0.5141	0.7792	0.7631	0.7792	0.7503
0.4233	26.0	104	0.5234	0.7833	0.7813	0.7833	0.7420
0.4233	27.0	108	0.5175	0.7833	0.7813	0.7833	0.7420
0.4233	28.0	112	0.5122	0.7958	0.7856	0.7958	0.7715
0.4233	29.0	116	0.5126	0.7958	0.7856	0.7958	0.7715
0.3931	30.0	120	0.5130	0.7958	0.7856	0.7958	0.7715

Framework versions

Transformers 4.33.3
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.13.3