README.md

metadata

license: apache-2.0
base_model: facebook/vit-msn-small
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: vit-msn-small-finetuned-alzheimers
    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: 0.996875

vit-msn-small-finetuned-alzheimers

This model is a fine-tuned version of facebook/vit-msn-small on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0160
• Accuracy: 0.9969

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• 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
0.2996	0.9778	22	0.3897	0.8438
0.3703	2.0	45	0.3595	0.8594
0.3087	2.9778	67	0.3777	0.8625
0.486	4.0	90	0.4530	0.8187
0.3307	4.9778	112	0.4560	0.8234
0.306	6.0	135	0.3471	0.8672
0.3005	6.9778	157	0.3025	0.8859
0.319	8.0	180	0.2451	0.8984
0.3489	8.9778	202	0.1814	0.9281
0.3251	10.0	225	0.2451	0.9156
0.3034	10.9778	247	0.1566	0.9406
0.2746	12.0	270	0.2493	0.8922
0.2369	12.9778	292	0.1622	0.9375
0.2231	14.0	315	0.1781	0.9359
0.2281	14.9778	337	0.1268	0.9531
0.2001	16.0	360	0.2431	0.9141
0.183	16.9778	382	0.1017	0.9625
0.1891	18.0	405	0.1802	0.9391
0.1862	18.9778	427	0.0869	0.9766
0.1935	20.0	450	0.1079	0.9688
0.1797	20.9778	472	0.1250	0.9563
0.1605	22.0	495	0.0655	0.9719
0.1848	22.9778	517	0.0806	0.9766
0.1498	24.0	540	0.1116	0.9578
0.1394	24.9778	562	0.0807	0.9672
0.1584	26.0	585	0.0525	0.9797
0.1302	26.9778	607	0.0513	0.9828
0.1356	28.0	630	0.0420	0.9875
0.1101	28.9778	652	0.0354	0.9875
0.1227	30.0	675	0.0583	0.9766
0.1158	30.9778	697	0.0253	0.9906
0.117	32.0	720	0.0231	0.9906
0.1022	32.9778	742	0.0726	0.9797
0.1221	34.0	765	0.0160	0.9969
0.0956	34.9778	787	0.0482	0.9844
0.0856	36.0	810	0.0256	0.9875
0.0996	36.9778	832	0.0211	0.9906
0.0848	38.0	855	0.0446	0.9797
0.1001	38.9778	877	0.0274	0.9875
0.0976	40.0	900	0.0225	0.9922
0.0864	40.9778	922	0.0207	0.9922
0.0865	42.0	945	0.0193	0.9969
0.0773	42.9778	967	0.0203	0.9922
0.075	44.0	990	0.0131	0.9969
0.0761	44.9778	1012	0.0129	0.9938
0.0624	46.0	1035	0.0114	0.9969
0.0557	46.9778	1057	0.0102	0.9953
0.0708	48.0	1080	0.0116	0.9953
0.0667	48.8889	1100	0.0131	0.9953

Framework versions

• Transformers 4.40.0
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1