vit-base-patch16-224-Trial006-YEL_STEM2

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: 0.0644
• 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
0.7114	1.0	2	0.6552	0.5962
0.6552	2.0	4	0.6270	0.6346
0.5889	3.0	6	0.5689	0.75
0.5458	4.0	8	0.5344	0.75
0.4979	5.0	10	0.4562	0.8462
0.4068	6.0	12	0.4825	0.7308
0.3409	7.0	14	0.3131	0.8846
0.3453	8.0	16	0.2443	0.9231
0.2641	9.0	18	0.2463	0.8654
0.2402	10.0	20	0.1693	0.9231
0.2718	11.0	22	0.2417	0.9038
0.2348	12.0	24	0.1242	0.9423
0.1613	13.0	26	0.1589	0.9231
0.1651	14.0	28	0.0644	1.0
0.1885	15.0	30	0.0554	1.0
0.1593	16.0	32	0.1013	0.9615
0.1424	17.0	34	0.2931	0.8846
0.1535	18.0	36	0.1800	0.9038
0.1699	19.0	38	0.1044	0.9423
0.1899	20.0	40	0.0576	0.9808
0.1469	21.0	42	0.1152	0.9615
0.1512	22.0	44	0.1046	0.9615
0.1215	23.0	46	0.1317	0.9231
0.1071	24.0	48	0.0721	0.9615
0.154	25.0	50	0.0647	0.9615
0.1505	26.0	52	0.1563	0.9615
0.1422	27.0	54	0.0764	0.9615
0.109	28.0	56	0.1009	0.9808
0.2124	29.0	58	0.0943	0.9615
0.1342	30.0	60	0.0885	0.9615
0.1063	31.0	62	0.1581	0.9615
0.104	32.0	64	0.0704	0.9615
0.1256	33.0	66	0.0539	0.9808
0.1365	34.0	68	0.0844	0.9615
0.1141	35.0	70	0.1329	0.9615
0.1326	36.0	72	0.0829	0.9615
0.1106	37.0	74	0.0655	0.9615
0.1178	38.0	76	0.0676	0.9615
0.1031	39.0	78	0.0653	0.9615
0.0991	40.0	80	0.0633	0.9615
0.0924	41.0	82	0.0416	0.9808
0.1009	42.0	84	0.0360	0.9808
0.0834	43.0	86	0.0334	0.9808
0.1132	44.0	88	0.0578	0.9615
0.1428	45.0	90	0.1010	0.9615
0.0962	46.0	92	0.1100	0.9615
0.0917	47.0	94	0.1021	0.9615
0.1292	48.0	96	0.0808	0.9615
0.1036	49.0	98	0.0625	0.9615
0.0986	50.0	100	0.0535	0.9615

Framework versions

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