metadata

license: apache-2.0
base_model: google/vit-base-patch16-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-base-patch16-224-U8-40d
    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: 0.8431372549019608

vit-base-patch16-224-U8-40d

This model is a fine-tuned version of google/vit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.6495
Accuracy: 0.8431

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: 6e-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: 0.05
num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3419	1.0	20	1.2998	0.4706
1.1313	2.0	40	1.0832	0.5686
0.7969	3.0	60	0.8094	0.6667
0.5063	4.0	80	0.6573	0.7843
0.3367	5.0	100	0.6389	0.7647
0.242	6.0	120	0.6879	0.7451
0.1881	7.0	140	0.7940	0.7059
0.1561	8.0	160	0.8030	0.7647
0.1557	9.0	180	0.7004	0.8235
0.1154	10.0	200	0.6495	0.8431
0.1469	11.0	220	1.1388	0.7059
0.0898	12.0	240	0.7967	0.7647
0.0719	13.0	260	0.8934	0.8039
0.0739	14.0	280	0.8476	0.7647
0.0823	15.0	300	0.9692	0.7647
0.0828	16.0	320	0.9385	0.7843
0.0761	17.0	340	1.1684	0.7255
0.0597	18.0	360	0.9414	0.7647
0.0727	19.0	380	1.0201	0.7059
0.0507	20.0	400	0.8563	0.8039
0.0587	21.0	420	0.8476	0.7843
0.0608	22.0	440	0.9399	0.8039
0.055	23.0	460	0.8820	0.7451
0.0619	24.0	480	1.0460	0.7647
0.0615	25.0	500	0.9392	0.8235
0.0455	26.0	520	0.9267	0.8235
0.0567	27.0	540	0.9784	0.7843
0.032	28.0	560	1.1541	0.7647
0.0276	29.0	580	0.8865	0.7843
0.0368	30.0	600	1.0848	0.8039
0.0342	31.0	620	0.9638	0.8039
0.037	32.0	640	0.9616	0.8039
0.0371	33.0	660	1.0073	0.8039
0.0371	34.0	680	1.0494	0.8039
0.0359	35.0	700	1.1287	0.7843
0.0255	36.0	720	1.1831	0.7647
0.0269	37.0	740	1.1610	0.7843
0.0292	38.0	760	1.1842	0.7843
0.0161	39.0	780	1.1092	0.8039
0.0333	40.0	800	1.1186	0.8039

Framework versions

Transformers 4.36.2
Pytorch 2.1.2+cu118
Datasets 2.16.1
Tokenizers 0.15.0