vit-base-patch16-224-Trial006_007_008-YEL_STEM3

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:

• Loss: 0.0514
• Accuracy: 1.0

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: 60
• eval_batch_size: 60
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 240
• 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
0.7488	0.8	3	0.7067	0.5253
0.6693	1.87	7	0.6195	0.6263
0.5816	2.93	11	0.5114	0.7273
0.4669	4.0	15	0.3586	0.8990
0.3998	4.8	18	0.2867	0.9192
0.3579	5.87	22	0.2326	0.9596
0.3087	6.93	26	0.1684	0.9596
0.2448	8.0	30	0.1525	0.9697
0.207	8.8	33	0.1863	0.9394
0.2283	9.87	37	0.1259	0.9697
0.2354	10.93	41	0.1090	0.9495
0.2452	12.0	45	0.0999	0.9596
0.2373	12.8	48	0.0702	0.9899
0.1516	13.87	52	0.0749	0.9697
0.1893	14.93	56	0.0808	0.9798
0.1761	16.0	60	0.0514	1.0
0.1971	16.8	63	0.0954	0.9798
0.1801	17.87	67	0.0536	0.9899
0.2004	18.93	71	0.0836	0.9596
0.1899	20.0	75	0.0500	0.9899
0.2222	20.8	78	0.0874	0.9798
0.1341	21.87	82	0.0371	1.0
0.2187	22.93	86	0.0545	1.0
0.1547	24.0	90	0.0445	1.0
0.1307	24.8	93	0.0404	1.0
0.1563	25.87	97	0.0377	1.0
0.1467	26.93	101	0.0405	1.0
0.1413	28.0	105	0.0379	1.0
0.1317	28.8	108	0.0367	1.0
0.1264	29.87	112	0.0359	1.0
0.128	30.93	116	0.0328	1.0
0.1292	32.0	120	0.0356	1.0
0.1549	32.8	123	0.0339	1.0
0.1272	33.87	127	0.0296	1.0
0.1376	34.93	131	0.0295	1.0
0.1884	36.0	135	0.0308	1.0
0.1536	36.8	138	0.0295	1.0
0.166	37.87	142	0.0299	1.0
0.0929	38.93	146	0.0307	1.0
0.1592	40.0	150	0.0307	1.0

Framework versions

• Transformers 4.30.0.dev0
• Pytorch 1.12.1
• Datasets 2.12.0
• Tokenizers 0.13.1