smids_5x_beit_base_adamax_0001_fold3

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.2034
• Accuracy: 0.88

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• 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
0.6864	1.0	375	0.6860	0.7217
0.4252	2.0	750	0.6416	0.7633
0.3187	3.0	1125	0.4319	0.845
0.2894	4.0	1500	0.4449	0.84
0.2471	5.0	1875	0.4205	0.8483
0.1902	6.0	2250	0.3865	0.8633
0.2206	7.0	2625	0.4140	0.86
0.1469	8.0	3000	0.4250	0.875
0.0865	9.0	3375	0.5396	0.8717
0.0868	10.0	3750	0.6873	0.8583
0.0906	11.0	4125	0.4484	0.8883
0.0585	12.0	4500	0.6803	0.88
0.1093	13.0	4875	0.5271	0.8833
0.0782	14.0	5250	0.6855	0.8633
0.0163	15.0	5625	0.6773	0.865
0.0996	16.0	6000	0.6530	0.8733
0.0415	17.0	6375	0.8051	0.88
0.0673	18.0	6750	0.7716	0.88
0.1076	19.0	7125	0.7362	0.8783
0.0103	20.0	7500	0.7231	0.875
0.0579	21.0	7875	0.7618	0.88
0.0461	22.0	8250	0.9110	0.8717
0.0544	23.0	8625	0.7656	0.8767
0.0075	24.0	9000	0.8172	0.8933
0.0242	25.0	9375	1.0276	0.865
0.001	26.0	9750	1.1126	0.8717
0.0272	27.0	10125	0.9522	0.8833
0.0034	28.0	10500	0.9353	0.87
0.0223	29.0	10875	0.8506	0.8833
0.0058	30.0	11250	1.1236	0.855
0.0185	31.0	11625	1.0029	0.8733
0.0001	32.0	12000	0.8488	0.88
0.0006	33.0	12375	0.8729	0.8767
0.0005	34.0	12750	1.0735	0.875
0.0003	35.0	13125	1.0286	0.885
0.0	36.0	13500	0.8268	0.89
0.0001	37.0	13875	1.0063	0.875
0.0002	38.0	14250	0.9833	0.885
0.0009	39.0	14625	0.9292	0.8917
0.0034	40.0	15000	0.9953	0.8883
0.0	41.0	15375	1.0555	0.885
0.0	42.0	15750	1.0377	0.89
0.0	43.0	16125	1.1991	0.8717
0.0	44.0	16500	1.1156	0.8783
0.0	45.0	16875	1.1077	0.8717
0.0001	46.0	17250	1.0635	0.8817
0.0	47.0	17625	1.1588	0.8817
0.0	48.0	18000	1.1268	0.8867
0.0	49.0	18375	1.2011	0.88
0.0	50.0	18750	1.2034	0.88

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2