vit-base-patch16-224-Trial007-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: 0.0269
• 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
0.8443	0.89	2	0.7813	0.3148
0.7501	1.78	4	0.7087	0.5556
0.6312	2.67	6	0.5306	0.9074
0.4329	4.0	9	0.3618	0.9074
0.4438	4.89	11	0.2699	0.9444
0.3858	5.78	13	0.3650	0.7963
0.339	6.67	15	0.1911	0.9630
0.2852	8.0	18	0.1611	0.9630
0.1866	8.89	20	0.1516	0.9444
0.1748	9.78	22	0.1333	0.9630
0.1996	10.67	24	0.1188	0.9630
0.1604	12.0	27	0.1169	0.9444
0.1319	12.89	29	0.0835	0.9815
0.141	13.78	31	0.0704	0.9815
0.123	14.67	33	0.0574	0.9815
0.0678	16.0	36	0.0604	0.9815
0.1208	16.89	38	0.0385	0.9815
0.0942	17.78	40	0.0269	1.0
0.0822	18.67	42	0.0169	1.0
0.0578	20.0	45	0.0175	1.0
0.0611	20.89	47	0.0220	1.0
0.1053	21.78	49	0.0098	1.0
0.1713	22.67	51	0.0156	1.0
0.0515	24.0	54	0.0111	1.0
0.1227	24.89	56	0.0166	1.0
0.0891	25.78	58	0.0093	1.0
0.0768	26.67	60	0.0090	1.0
0.0755	28.0	63	0.0108	1.0
0.0798	28.89	65	0.0201	1.0
0.1005	29.78	67	0.0118	1.0
0.1113	30.67	69	0.0131	1.0
0.1034	32.0	72	0.0171	1.0
0.0857	32.89	74	0.0158	1.0
0.0864	33.78	76	0.0141	1.0
0.1241	34.67	78	0.0127	1.0
0.0868	36.0	81	0.0118	1.0
0.0704	36.89	83	0.0113	1.0
0.0938	37.78	85	0.0109	1.0
0.1181	38.67	87	0.0120	1.0
0.0509	40.0	90	0.0149	1.0
0.0684	40.89	92	0.0155	1.0
0.0625	41.78	94	0.0151	1.0
0.0746	42.67	96	0.0143	1.0
0.1062	44.0	99	0.0133	1.0
0.0579	44.44	100	0.0132	1.0

Framework versions

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