metadata

library_name: transformers
license: apache-2.0
base_model: google/vit-base-patch32-384
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-base-patch32-384-finetuned-eurosat-albumentations
    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: 0.9726027397260274

vit-base-patch32-384-finetuned-eurosat-albumentations

This model is a fine-tuned version of google/vit-base-patch32-384 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.1871
Accuracy: 0.9726

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: 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_ratio: 0.1
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7204	0.9412	12	0.5695	0.7397
0.4269	1.9804	25	0.2537	0.9178
0.1605	2.9412	37	0.3347	0.8767
0.0758	3.9804	50	0.2203	0.9041
0.0405	4.9412	62	0.3563	0.9178
0.0358	5.9804	75	0.2326	0.9315
0.0188	6.9412	87	0.2046	0.9315
0.026	7.9804	100	0.2195	0.8904
0.0582	8.9412	112	0.3378	0.9178
0.0113	9.9804	125	0.2685	0.9178
0.0081	10.9412	137	0.2443	0.9315
0.0091	11.9804	150	0.4675	0.9041
0.0065	12.9412	162	0.3252	0.9452
0.0026	13.9804	175	0.1871	0.9726
0.0043	14.9412	187	0.2256	0.9589
0.0094	15.9804	200	0.1980	0.9452
0.0028	16.9412	212	0.2928	0.9315
0.0003	17.9804	225	0.2241	0.9726
0.0006	18.9412	237	0.2396	0.9726
0.0012	19.9804	250	0.2663	0.9315
0.0001	20.9412	262	0.2266	0.9726
0.0002	21.9804	275	0.2637	0.9452
0.0001	22.9412	287	0.2873	0.9452
0.0003	23.9804	300	0.2068	0.9589
0.0001	24.9412	312	0.2485	0.9452
0.0047	25.9804	325	0.3375	0.9178
0.0015	26.9412	337	0.3132	0.9589
0.0001	27.9804	350	0.3148	0.9452
0.0025	28.9412	362	0.2533	0.9452
0.0038	29.9804	375	0.2860	0.9315
0.0025	30.9412	387	0.2785	0.9452
0.0031	31.9804	400	0.3246	0.9452
0.0	32.9412	412	0.3367	0.9452
0.0006	33.9804	425	0.2625	0.9726
0.0	34.9412	437	0.2689	0.9589
0.0007	35.9804	450	0.2891	0.9726
0.0003	36.9412	462	0.4523	0.9315
0.0003	37.9804	475	0.3426	0.9452
0.0001	38.9412	487	0.3167	0.9589
0.0	39.9804	500	0.3237	0.9589
0.0002	40.9412	512	0.3085	0.9589
0.0	41.9804	525	0.3095	0.9589
0.0	42.9412	537	0.3049	0.9589
0.0002	43.9804	550	0.3039	0.9589
0.0001	44.9412	562	0.3044	0.9589
0.0001	45.9804	575	0.3031	0.9726
0.0	46.9412	587	0.3028	0.9726
0.0	47.9804	600	0.3027	0.9726

Framework versions

Transformers 4.46.2
Pytorch 2.5.0+cu121
Datasets 2.21.0
Tokenizers 0.20.3