vit-base-patch16-224-Trial006-YEL_STEM1

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: 0.0494
• 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
0.7481	0.89	2	0.6901	0.5455
0.6788	1.78	4	0.6743	0.6364
0.71	2.67	6	0.5935	0.6909
0.5911	4.0	9	0.5331	0.7091
0.572	4.89	11	0.4855	0.7636
0.5206	5.78	13	0.5816	0.7273
0.5029	6.67	15	0.4695	0.7818
0.4806	8.0	18	0.4680	0.8
0.3945	8.89	20	0.4059	0.8364
0.356	9.78	22	0.3764	0.8727
0.3153	10.67	24	0.3162	0.9091
0.3563	12.0	27	0.2654	0.8909
0.2904	12.89	29	0.2471	0.9091
0.2425	13.78	31	0.2265	0.8909
0.2402	14.67	33	0.2225	0.8909
0.2309	16.0	36	0.1752	0.9273
0.294	16.89	38	0.1769	0.9273
0.2004	17.78	40	0.1878	0.9091
0.2229	18.67	42	0.1126	0.9455
0.2804	20.0	45	0.1212	0.9273
0.2113	20.89	47	0.0494	1.0
0.1744	21.78	49	0.0767	0.9818
0.1645	22.67	51	0.0531	0.9818
0.2322	24.0	54	0.0757	0.9455
0.1934	24.89	56	0.0302	1.0
0.2172	25.78	58	0.1479	0.9455
0.1918	26.67	60	0.0374	0.9818
0.1948	28.0	63	0.1695	0.9273
0.2099	28.89	65	0.0743	0.9818
0.155	29.78	67	0.0275	1.0
0.1563	30.67	69	0.0273	1.0
0.149	32.0	72	0.0628	0.9818
0.1767	32.89	74	0.0388	0.9818
0.1828	33.78	76	0.0289	1.0
0.1825	34.67	78	0.0522	0.9636
0.197	36.0	81	0.0187	1.0
0.1534	36.89	83	0.0200	1.0
0.2099	37.78	85	0.0269	0.9818
0.1694	38.67	87	0.0176	1.0
0.101	40.0	90	0.0140	1.0
0.1488	40.89	92	0.0162	1.0
0.184	41.78	94	0.0175	1.0
0.182	42.67	96	0.0148	1.0
0.1364	44.0	99	0.0139	1.0
0.1332	44.44	100	0.0137	1.0

Framework versions

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