README.md

metadata

license: apache-2.0
base_model: microsoft/swin-tiny-patch4-window7-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: swin-tiny-patch4-window7-224-finetuned-eurosat
    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: 0.8803088803088803

swin-tiny-patch4-window7-224-finetuned-eurosat

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: 0.5056
• Accuracy: 0.8803

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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.1952	0.99	18	1.5914	0.5985
1.3705	1.97	36	1.2164	0.6873
1.026	2.96	54	0.9974	0.7375
0.829	4.0	73	0.7667	0.7722
0.6513	4.99	91	0.6674	0.8224
0.5516	5.97	109	0.5810	0.8378
0.4978	6.96	127	0.5498	0.8263
0.4568	8.0	146	0.5999	0.8185
0.4047	8.99	164	0.5211	0.8494
0.3696	9.97	182	0.5201	0.8571
0.3479	10.96	200	0.5310	0.8263
0.329	12.0	219	0.5439	0.8494
0.3376	12.99	237	0.5050	0.8494
0.2804	13.97	255	0.5709	0.8263
0.2941	14.96	273	0.6376	0.8147
0.3026	16.0	292	0.5447	0.8494
0.2578	16.99	310	0.5056	0.8803
0.219	17.97	328	0.5620	0.8610
0.2403	18.96	346	0.5582	0.8456
0.2258	20.0	365	0.5458	0.8494
0.2265	20.99	383	0.5411	0.8533
0.1893	21.97	401	0.5477	0.8494
0.1896	22.96	419	0.5125	0.8494
0.1976	24.0	438	0.5672	0.8340
0.1725	24.99	456	0.5581	0.8456
0.168	25.97	474	0.5965	0.8456
0.1821	26.96	492	0.5567	0.8610
0.1805	28.0	511	0.5998	0.8533
0.1616	28.99	529	0.5451	0.8533
0.1467	29.97	547	0.5574	0.8494
0.1439	30.96	565	0.5707	0.8571
0.13	32.0	584	0.6019	0.8378
0.1353	32.99	602	0.5952	0.8610
0.1329	33.97	620	0.6262	0.8378
0.1258	34.96	638	0.6314	0.8456
0.1408	36.0	657	0.5761	0.8494
0.1197	36.99	675	0.5703	0.8610
0.1208	37.97	693	0.6247	0.8456
0.1197	38.96	711	0.6026	0.8533
0.1271	40.0	730	0.5953	0.8533
0.1053	40.99	748	0.6070	0.8533
0.0846	41.97	766	0.6094	0.8610
0.1206	42.96	784	0.5912	0.8494
0.1225	44.0	803	0.6074	0.8494
0.1184	44.99	821	0.5943	0.8494
0.1027	45.97	839	0.6084	0.8494
0.1113	46.96	857	0.6034	0.8533
0.0945	48.0	876	0.6106	0.8494
0.1159	48.99	894	0.6143	0.8533
0.0963	49.32	900	0.6144	0.8533

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2