metadata

license: apache-2.0
tags:
  - image-classification
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: raildefectfft1
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: defect
          type: imagefolder
          config: Dhika--defectfft
          split: validation
          args: Dhika--defectfft
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.7914285714285715

raildefectfft1

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the defect dataset. It achieves the following results on the evaluation set:

Loss: 0.7259
Accuracy: 0.7914

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: 0.0002
train_batch_size: 30
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3927	0.67	10	1.1308	0.6429
0.8111	1.33	20	0.9788	0.6629
0.513	2.0	30	0.7938	0.74
0.2943	2.67	40	0.8517	0.7343
0.2029	3.33	50	0.7300	0.7686
0.1629	4.0	60	0.7259	0.7914
0.1131	4.67	70	0.9103	0.7314
0.0955	5.33	80	0.8504	0.7657
0.0547	6.0	90	1.0702	0.72
0.0489	6.67	100	1.1708	0.6971
0.0382	7.33	110	1.2376	0.6943
0.0356	8.0	120	1.3361	0.6857
0.0311	8.67	130	1.1809	0.7229
0.0346	9.33	140	1.3405	0.7086
0.0378	10.0	150	1.1800	0.7171
0.0326	10.67	160	1.1292	0.7343
0.0319	11.33	170	1.0885	0.7371
0.0347	12.0	180	1.4550	0.6771
0.0283	12.67	190	1.1957	0.7314
0.0336	13.33	200	1.4648	0.6743
0.0175	14.0	210	1.4927	0.6771
0.0167	14.67	220	1.3760	0.7057
0.0149	15.33	230	1.2464	0.7229
0.0154	16.0	240	1.2553	0.7257
0.0135	16.67	250	1.2768	0.7314
0.0133	17.33	260	1.2857	0.7343
0.0122	18.0	270	1.2905	0.7314
0.0121	18.67	280	1.2929	0.7314
0.0115	19.33	290	1.2958	0.7314
0.0111	20.0	300	1.2985	0.7314
0.011	20.67	310	1.3020	0.7343
0.0103	21.33	320	1.3051	0.7371
0.0103	22.0	330	1.3075	0.7371
0.0104	22.67	340	1.3098	0.7371
0.0096	23.33	350	1.3128	0.7371
0.0095	24.0	360	1.3154	0.7371
0.0096	24.67	370	1.3162	0.7371
0.0093	25.33	380	1.3183	0.7371
0.0091	26.0	390	1.3200	0.7371
0.0092	26.67	400	1.3213	0.7371
0.0089	27.33	410	1.3219	0.7371
0.0092	28.0	420	1.3224	0.7371
0.0089	28.67	430	1.3228	0.7371
0.0089	29.33	440	1.3231	0.7371
0.0089	30.0	450	1.3233	0.7371

Framework versions

Transformers 4.30.1
Pytorch 2.0.1+cu118
Datasets 2.12.0
Tokenizers 0.13.3