vit-base-patch16-224-Trial007-YEL_STEM

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: 0.0373
• 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
0.7081	0.89	2	0.6818	0.5556
0.6584	1.78	4	0.5915	0.7037
0.5552	2.67	6	0.5366	0.7407
0.3763	4.0	9	0.3560	0.8519
0.397	4.89	11	0.2999	0.8519
0.3313	5.78	13	0.2307	0.9074
0.2957	6.67	15	0.1746	0.9259
0.2383	8.0	18	0.1432	0.9444
0.2664	8.89	20	0.3320	0.9074
0.2242	9.78	22	0.1120	0.9630
0.2072	10.67	24	0.0718	0.9630
0.1399	12.0	27	0.0494	0.9815
0.1846	12.89	29	0.0373	1.0
0.1816	13.78	31	0.0354	1.0
0.1453	14.67	33	0.0461	0.9815
0.1406	16.0	36	0.0333	1.0
0.1749	16.89	38	0.0275	1.0
0.1383	17.78	40	0.0203	1.0
0.1659	18.67	42	0.0186	1.0
0.153	20.0	45	0.0184	1.0
0.1497	20.89	47	0.0215	1.0
0.1088	21.78	49	0.0408	0.9815
0.1796	22.67	51	0.0377	0.9815
0.1041	24.0	54	0.0631	0.9815
0.1193	24.89	56	0.0637	0.9815
0.1653	25.78	58	0.0730	0.9815
0.1296	26.67	60	0.0779	0.9815
0.1036	28.0	63	0.0312	0.9815
0.1287	28.89	65	0.0116	1.0
0.1307	29.78	67	0.0129	1.0
0.1337	30.67	69	0.0141	1.0
0.1274	32.0	72	0.0161	1.0
0.1612	32.89	74	0.0177	1.0
0.1504	33.78	76	0.0181	1.0
0.1307	34.67	78	0.0175	1.0
0.125	36.0	81	0.0170	1.0
0.1357	36.89	83	0.0165	1.0
0.1033	37.78	85	0.0162	1.0
0.1749	38.67	87	0.0164	1.0
0.0906	40.0	90	0.0153	1.0
0.1349	40.89	92	0.0152	1.0
0.1056	41.78	94	0.0150	1.0
0.1328	42.67	96	0.0148	1.0
0.0742	44.0	99	0.0148	1.0
0.0875	44.44	100	0.0148	1.0

Framework versions

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