vit-base-patch16-224-Trial007-YEL_STEM4

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: 0.1948
• 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
0.8588	0.89	2	0.7925	0.4815
0.7235	1.78	4	0.6471	0.6852
0.6009	2.67	6	0.5246	0.7222
0.4196	4.0	9	0.3422	0.9074
0.4022	4.89	11	0.3213	0.9259
0.3531	5.78	13	0.1948	1.0
0.3095	6.67	15	0.1196	1.0
0.283	8.0	18	0.0666	1.0
0.1607	8.89	20	0.0401	1.0
0.1459	9.78	22	0.0302	1.0
0.1325	10.67	24	0.0223	1.0
0.1362	12.0	27	0.0205	1.0
0.1623	12.89	29	0.0094	1.0
0.0974	13.78	31	0.0046	1.0
0.1077	14.67	33	0.0054	1.0
0.0742	16.0	36	0.0040	1.0
0.1468	16.89	38	0.0030	1.0
0.077	17.78	40	0.0041	1.0
0.0907	18.67	42	0.0109	1.0
0.0363	20.0	45	0.0023	1.0
0.0519	20.89	47	0.0016	1.0
0.0672	21.78	49	0.0015	1.0
0.0894	22.67	51	0.0020	1.0
0.0267	24.0	54	0.0020	1.0
0.0639	24.89	56	0.0019	1.0
0.0675	25.78	58	0.0023	1.0
0.0508	26.67	60	0.0020	1.0
0.0509	28.0	63	0.0014	1.0
0.0573	28.89	65	0.0018	1.0
0.0584	29.78	67	0.0014	1.0
0.0657	30.67	69	0.0012	1.0
0.0635	32.0	72	0.0009	1.0
0.0617	32.89	74	0.0008	1.0
0.0614	33.78	76	0.0008	1.0
0.0614	34.67	78	0.0009	1.0
0.0618	36.0	81	0.0008	1.0
0.0384	36.89	83	0.0008	1.0
0.0565	37.78	85	0.0008	1.0
0.0784	38.67	87	0.0008	1.0
0.0313	40.0	90	0.0007	1.0
0.0496	40.89	92	0.0007	1.0
0.0273	41.78	94	0.0008	1.0
0.0448	42.67	96	0.0008	1.0
0.0948	44.0	99	0.0007	1.0
0.0371	44.44	100	0.0007	1.0

Framework versions

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