metadata

license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
  - image-classification
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-weld-classify
    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: 0.6894977168949772

vit-weld-classify

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

Loss: 0.7966
Accuracy: 0.6895

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: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 18
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8686	0.8130	100	0.7966	0.6895
0.6935	1.6260	200	1.2217	0.5068
0.4225	2.4390	300	0.9592	0.6210
0.2586	3.2520	400	1.3123	0.5936
0.237	4.0650	500	0.8075	0.6986
0.2658	4.8780	600	1.0878	0.6210
0.1904	5.6911	700	1.1048	0.7169
0.0964	6.5041	800	1.3602	0.6849
0.0474	7.3171	900	1.1331	0.7671
0.1179	8.1301	1000	1.1228	0.7306
0.0447	8.9431	1100	1.2609	0.7397
0.0043	9.7561	1200	1.1746	0.7763
0.1059	10.5691	1300	1.1867	0.7763
0.0026	11.3821	1400	1.2890	0.7534
0.0039	12.1951	1500	1.3283	0.7580
0.002	13.0081	1600	1.1871	0.7671
0.0019	13.8211	1700	1.1643	0.7900
0.0264	14.6341	1800	1.1537	0.7900
0.0015	15.4472	1900	1.1821	0.7945
0.0015	16.2602	2000	1.1962	0.7900
0.0014	17.0732	2100	1.2036	0.7900
0.0014	17.8862	2200	1.2067	0.7900

Framework versions

Transformers 4.41.1
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1