vit-msn-small-corect_dataset_lateral_flow_ivalidation

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: 0.2930
• Accuracy: 0.9048

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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9231	3	0.6350	0.6337
No log	1.8462	6	0.5047	0.8022
No log	2.7692	9	0.3701	0.8791
0.5485	4.0	13	0.5379	0.7436
0.5485	4.9231	16	0.2748	0.8938
0.5485	5.8462	19	0.3004	0.8974
0.3335	6.7692	22	0.3492	0.8681
0.3335	8.0	26	0.2497	0.8974
0.3335	8.9231	29	0.4304	0.8315
0.3087	9.8462	32	0.3479	0.8791
0.3087	10.7692	35	0.3796	0.8645
0.3087	12.0	39	0.4152	0.8352
0.2614	12.9231	42	0.3199	0.9011
0.2614	13.8462	45	0.3434	0.8718
0.2614	14.7692	48	0.4001	0.8462
0.2471	16.0	52	0.3220	0.8901
0.2471	16.9231	55	0.3540	0.8718
0.2471	17.8462	58	0.4019	0.8535
0.2817	18.7692	61	0.3152	0.8974
0.2817	20.0	65	0.3978	0.8571
0.2817	20.9231	68	0.4289	0.8388
0.2353	21.8462	71	0.3146	0.8974
0.2353	22.7692	74	0.3206	0.8864
0.2353	24.0	78	0.3715	0.8828
0.2339	24.9231	81	0.3446	0.8938
0.2339	25.8462	84	0.2930	0.9048
0.2339	26.7692	87	0.4349	0.8205
0.2301	28.0	91	0.3630	0.8681
0.2301	28.9231	94	0.3669	0.8645
0.2301	29.8462	97	0.5037	0.7912
0.2115	30.7692	100	0.3449	0.8828
0.2115	32.0	104	0.3280	0.9011
0.2115	32.9231	107	0.4031	0.8425
0.2033	33.8462	110	0.3612	0.8535
0.2033	34.7692	113	0.3163	0.8901
0.2033	36.0	117	0.3234	0.8864
0.1807	36.9231	120	0.3307	0.8791

Framework versions

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