vit-base-patch16-224-Trial006-007-008-YEL_STEM3

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: 0.1618
• Accuracy: 0.9241

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
0.7035	1.0	6	0.6557	0.6203
0.6168	2.0	12	0.5788	0.7215
0.5412	3.0	18	0.5005	0.7785
0.496	4.0	24	0.4946	0.7722
0.4024	5.0	30	0.4057	0.8165
0.4098	6.0	36	0.3076	0.8544
0.3645	7.0	42	0.3250	0.8418
0.276	8.0	48	0.2206	0.8924
0.3358	9.0	54	0.2100	0.8987
0.3386	10.0	60	0.1618	0.9241
0.2778	11.0	66	0.1609	0.9177
0.25	12.0	72	0.1581	0.9114
0.2914	13.0	78	0.1663	0.9114
0.2273	14.0	84	0.1525	0.9177
0.2694	15.0	90	0.1708	0.9051
0.2745	16.0	96	0.2364	0.8734
0.2809	17.0	102	0.1976	0.8608
0.2368	18.0	108	0.1517	0.9114
0.328	19.0	114	0.2454	0.8671
0.2571	20.0	120	0.1482	0.9114
0.2996	21.0	126	0.1629	0.8987
0.266	22.0	132	0.1360	0.9114
0.2323	23.0	138	0.1427	0.9114
0.2285	24.0	144	0.1683	0.9051
0.2566	25.0	150	0.1442	0.9114
0.2509	26.0	156	0.1595	0.9114
0.2337	27.0	162	0.1291	0.9177
0.2203	28.0	168	0.1302	0.8987
0.2409	29.0	174	0.1274	0.9114
0.2256	30.0	180	0.1272	0.8987
0.2157	31.0	186	0.1289	0.9177
0.2168	32.0	192	0.1267	0.9114
0.2426	33.0	198	0.1438	0.8987
0.2404	34.0	204	0.1388	0.8987
0.2218	35.0	210	0.1243	0.9241
0.3068	36.0	216	0.1268	0.9241
0.1721	37.0	222	0.1477	0.8987
0.2201	38.0	228	0.1545	0.8987
0.2581	39.0	234	0.1700	0.8987
0.213	40.0	240	0.1254	0.9114
0.2953	41.0	246	0.1237	0.9114
0.2564	42.0	252	0.1472	0.9051
0.249	43.0	258	0.1409	0.9051
0.2372	44.0	264	0.1495	0.9114
0.2541	45.0	270	0.1412	0.9051
0.1997	46.0	276	0.1308	0.9114
0.2381	47.0	282	0.1253	0.9177
0.2623	48.0	288	0.1267	0.9051
0.1855	49.0	294	0.1285	0.9051
0.1877	50.0	300	0.1289	0.9051

Framework versions

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