google-vit-base-patch16-224-Waste-O-I-classification

This model is a fine-tuned version of google/vit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Accuracy: 0.956
• Loss: 0.3036

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
0.2168	0.1580	1000	0.9525	0.1303
0.196	0.3159	2000	0.941	0.1638
0.1993	0.4739	3000	0.9285	0.2206
0.1849	0.6318	4000	0.9225	0.2288
0.199	0.7898	5000	0.9105	0.3331
0.2171	0.9477	6000	0.944	0.1582
0.1209	1.1057	7000	0.9495	0.1887
0.114	1.2636	8000	0.932	0.1950
0.1268	1.4216	9000	0.9335	0.1965
0.1272	1.5795	10000	0.9165	0.3112
0.1003	1.7375	11000	0.9575	0.1353
0.0844	1.8954	12000	0.9345	0.2635
0.0757	2.0534	13000	0.952	0.1434
0.053	2.2113	14000	0.933	0.3203
0.0994	2.3693	15000	0.9405	0.2165
0.0248	2.5272	16000	0.951	0.2400
0.0842	2.6852	17000	0.906	0.4092
0.0733	2.8432	18000	0.9515	0.1937
0.0542	3.0011	19000	0.938	0.2911
0.0202	3.1591	20000	0.936	0.3648
0.0237	3.3170	21000	0.9355	0.3618
0.0294	3.4750	22000	0.9255	0.4209
0.0375	3.6329	23000	0.943	0.2840
0.0176	3.7909	24000	0.9525	0.2604
0.0252	3.9488	25000	0.9515	0.2500
0.0024	4.1068	26000	0.9545	0.2892
0.0119	4.2647	27000	0.956	0.3036

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0+cpu
• Datasets 2.20.0
• Tokenizers 0.19.1