metadata

library_name: transformers
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-weldclassifyv4
    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: 0.920863309352518

vit-weldclassifyv4

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: 0.3301
Accuracy: 0.9209

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: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8207	0.6410	100	1.0336	0.5647
0.6506	1.2821	200	1.1982	0.5791
0.5324	1.9231	300	0.6060	0.7770
0.2486	2.5641	400	0.7294	0.7518
0.1366	3.2051	500	0.4832	0.8417
0.3124	3.8462	600	0.8676	0.7626
0.0296	4.4872	700	0.4233	0.8885
0.0723	5.1282	800	0.4470	0.8849
0.0342	5.7692	900	0.3406	0.9173
0.0055	6.4103	1000	0.3301	0.9209
0.0048	7.0513	1100	0.3471	0.9173
0.0036	7.6923	1200	0.3346	0.9137
0.003	8.3333	1300	0.3498	0.9137
0.003	8.9744	1400	0.3549	0.9101
0.0027	9.6154	1500	0.3569	0.9137

Framework versions

Transformers 4.44.2
Pytorch 2.5.0+cu121
Datasets 3.1.0
Tokenizers 0.19.1