README.md

metadata

library_name: transformers
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-base-patch16-224-in21k-finetuned-lf-invalidation
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: test
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.951063829787234

vit-base-patch16-224-in21k-finetuned-lf-invalidation

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: 0.1798
• Accuracy: 0.9511

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6773	0.9796	12	0.6550	0.5681
0.5982	1.9592	24	0.5839	0.6362
0.479	2.9388	36	0.4356	0.8894
0.3862	4.0	49	0.2807	0.9362
0.2498	4.9796	61	0.2599	0.9128
0.2836	5.9592	73	0.5015	0.7745
0.2641	6.9388	85	0.5500	0.7340
0.2716	8.0	98	0.3083	0.8787
0.2382	8.9796	110	0.2885	0.8936
0.1985	9.9592	122	0.1798	0.9511
0.2174	10.9388	134	0.3060	0.8766
0.2372	12.0	147	0.3084	0.8702
0.2164	12.9796	159	0.2667	0.9021
0.2106	13.9592	171	0.3747	0.8447
0.1956	14.9388	183	0.5105	0.7851
0.2154	16.0	196	0.5683	0.7787
0.179	16.9796	208	0.4279	0.8340
0.2548	17.9592	220	0.6493	0.7404
0.236	18.9388	232	0.3860	0.8340
0.2121	20.0	245	0.5826	0.7766
0.1691	20.9796	257	0.3195	0.8638
0.1824	21.9592	269	0.3772	0.8404
0.1733	22.9388	281	0.5182	0.7936
0.1837	24.0	294	0.4924	0.8149
0.1274	24.9796	306	0.3895	0.8447
0.1415	25.9592	318	0.3662	0.8532
0.186	26.9388	330	0.4347	0.8447
0.1403	28.0	343	0.4490	0.8383
0.1635	28.9796	355	0.7771	0.7085
0.2135	29.9592	367	0.3503	0.8702
0.1456	30.9388	379	0.3815	0.8617
0.1634	32.0	392	0.2810	0.9
0.1308	32.9796	404	0.4643	0.8383
0.163	33.9592	416	0.3337	0.8787
0.1736	34.9388	428	0.4070	0.8553
0.1638	36.0	441	0.4142	0.8574
0.1488	36.9796	453	0.5039	0.8170
0.148	37.9592	465	0.5767	0.7745
0.1741	38.9388	477	0.4842	0.8255
0.1338	40.0	490	0.7236	0.7234
0.1302	40.9796	502	0.5295	0.8043
0.141	41.9592	514	0.5294	0.8085
0.1461	42.9388	526	0.5485	0.7979
0.1006	44.0	539	0.5453	0.7915
0.1317	44.9796	551	0.5930	0.7681
0.1069	45.9592	563	0.4976	0.8170
0.1531	46.9388	575	0.5105	0.8064
0.155	48.0	588	0.6128	0.7638
0.1237	48.9796	600	0.6180	0.7617

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1