vit-msn-small-finetuned-lf-invalidation

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: 0.2442
• Accuracy: 0.9234

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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
No log	0.96	6	0.7163	0.4
0.7	1.92	12	0.4886	0.8234
0.7	2.88	18	0.4683	0.7596
0.3793	4.0	25	0.3421	0.8447
0.383	4.96	31	0.2535	0.9191
0.383	5.92	37	0.2442	0.9234
0.3658	6.88	43	0.3795	0.8404
0.2601	8.0	50	0.3831	0.8383
0.2601	8.96	56	0.3993	0.8191
0.26	9.92	62	0.2265	0.8979
0.26	10.88	68	0.5355	0.7319
0.3376	12.0	75	0.3881	0.8043
0.248	12.96	81	0.2618	0.8979
0.248	13.92	87	0.5545	0.7362
0.2133	14.88	93	0.9307	0.5489
0.2576	16.0	100	0.4236	0.8149
0.2576	16.96	106	0.4333	0.8106
0.2466	17.92	112	0.8464	0.6128
0.2466	18.88	118	0.7970	0.6489
0.228	20.0	125	0.3522	0.8660
0.2542	20.96	131	0.5095	0.7872
0.2542	21.92	137	0.4808	0.8021
0.2032	22.88	143	0.5805	0.7340
0.1998	24.0	150	0.3987	0.8319
0.1998	24.96	156	0.4889	0.7894
0.1565	25.92	162	0.8003	0.6468
0.1565	26.88	168	0.4740	0.7936
0.1934	28.0	175	0.4442	0.8319
0.1878	28.96	181	0.7115	0.7021
0.1878	29.92	187	0.4234	0.8277
0.1848	30.88	193	0.6975	0.6957
0.1705	32.0	200	0.2965	0.8894
0.1705	32.96	206	0.8020	0.6766
0.1744	33.92	212	0.7330	0.6979
0.1744	34.88	218	1.0977	0.5723
0.1707	36.0	225	1.0648	0.5894
0.1719	36.96	231	0.8495	0.6404
0.1719	37.92	237	0.3177	0.8787
0.1839	38.88	243	0.4839	0.7894
0.1544	40.0	250	0.4100	0.8362
0.1544	40.96	256	0.6012	0.7553
0.135	41.92	262	0.6832	0.7213
0.135	42.88	268	0.6663	0.7170
0.14	44.0	275	0.6219	0.7383
0.151	44.96	281	0.9176	0.6149
0.151	45.92	287	0.8830	0.6404
0.1284	46.88	293	0.7249	0.7043
0.1586	48.0	300	0.7146	0.7043

Framework versions

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