vit-base-patch16-224-Trial006-007-008-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: 0.1837
• Accuracy: 0.9492

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
0.7399	0.89	4	0.7281	0.5085
0.6861	2.0	9	0.6326	0.6186
0.6362	2.89	13	0.6392	0.6610
0.5728	4.0	18	0.5497	0.6695
0.515	4.89	22	0.4464	0.8136
0.5437	6.0	27	0.4090	0.8136
0.4647	6.89	31	0.3873	0.8220
0.4616	8.0	36	0.4446	0.7712
0.4396	8.89	40	0.4029	0.8220
0.4354	10.0	45	0.3104	0.8559
0.4187	10.89	49	0.4200	0.8220
0.4639	12.0	54	0.2883	0.8898
0.3663	12.89	58	0.2667	0.9153
0.3169	14.0	63	0.3731	0.8644
0.4725	14.89	67	0.2759	0.8729
0.4438	16.0	72	0.3782	0.8559
0.361	16.89	76	0.2790	0.8983
0.3321	18.0	81	0.2890	0.8983
0.3071	18.89	85	0.2499	0.8814
0.333	20.0	90	0.2271	0.9068
0.2706	20.89	94	0.2631	0.8814
0.2963	22.0	99	0.2449	0.9068
0.323	22.89	103	0.1904	0.9322
0.2677	24.0	108	0.2341	0.9237
0.2473	24.89	112	0.2191	0.9237
0.2598	26.0	117	0.2106	0.9322
0.2733	26.89	121	0.1837	0.9492
0.2506	28.0	126	0.1828	0.9407
0.2315	28.89	130	0.2110	0.9153
0.2519	30.0	135	0.2288	0.9153
0.289	30.89	139	0.1781	0.9322
0.3309	32.0	144	0.1571	0.9322
0.2435	32.89	148	0.1778	0.9322
0.2643	34.0	153	0.2275	0.9153
0.2238	34.89	157	0.1777	0.9322
0.2773	36.0	162	0.1955	0.9322
0.2422	36.89	166	0.2019	0.9237
0.2812	38.0	171	0.1794	0.9492
0.2626	38.89	175	0.1884	0.9407
0.2059	40.0	180	0.1843	0.9322
0.2585	40.89	184	0.1721	0.9492
0.2022	42.0	189	0.1640	0.9492
0.2403	42.89	193	0.1660	0.9407
0.2771	44.0	198	0.1686	0.9407
0.2153	44.44	200	0.1684	0.9407

Framework versions

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