smids_10x_beit_large_adamax_001_fold4

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

• Loss: 1.6842
• Accuracy: 0.8717

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: 0.001
• 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
0.3361	1.0	750	0.4333	0.8367
0.2968	2.0	1500	0.4495	0.8467
0.288	3.0	2250	0.4264	0.8383
0.2379	4.0	3000	0.4907	0.85
0.1893	5.0	3750	0.4876	0.8533
0.1419	6.0	4500	0.4376	0.8667
0.1288	7.0	5250	0.5742	0.84
0.079	8.0	6000	0.6426	0.86
0.0885	9.0	6750	0.6694	0.8617
0.0513	10.0	7500	0.7772	0.8483
0.0371	11.0	8250	0.7425	0.8667
0.0559	12.0	9000	0.7844	0.8633
0.0437	13.0	9750	0.9475	0.8617
0.0237	14.0	10500	0.8539	0.86
0.0064	15.0	11250	1.1662	0.8683
0.0766	16.0	12000	1.1003	0.8683
0.0045	17.0	12750	1.1294	0.8633
0.0012	18.0	13500	1.0595	0.8717
0.0107	19.0	14250	1.0246	0.875
0.0098	20.0	15000	0.9670	0.8633
0.0227	21.0	15750	1.0829	0.8633
0.0004	22.0	16500	1.0091	0.855
0.0026	23.0	17250	1.0123	0.8667
0.001	24.0	18000	1.0183	0.8783
0.0083	25.0	18750	1.2133	0.8533
0.0076	26.0	19500	1.0638	0.865
0.0045	27.0	20250	1.1546	0.8717
0.0001	28.0	21000	1.0902	0.8567
0.0003	29.0	21750	1.1809	0.86
0.0	30.0	22500	1.2715	0.8733
0.0001	31.0	23250	1.1922	0.8767
0.0	32.0	24000	1.4076	0.87
0.0075	33.0	24750	1.3961	0.8617
0.0	34.0	25500	1.4345	0.875
0.0	35.0	26250	1.6125	0.8683
0.0	36.0	27000	1.5456	0.8567
0.0	37.0	27750	1.5632	0.865
0.0	38.0	28500	1.6349	0.8617
0.0	39.0	29250	1.5362	0.8617
0.0	40.0	30000	1.6434	0.8667
0.0	41.0	30750	1.6815	0.87
0.0	42.0	31500	1.6593	0.8667
0.0	43.0	32250	1.6757	0.87
0.0	44.0	33000	1.6503	0.8683
0.0	45.0	33750	1.6999	0.8667
0.0	46.0	34500	1.6868	0.8667
0.0	47.0	35250	1.6803	0.87
0.0	48.0	36000	1.6872	0.8733
0.0	49.0	36750	1.6911	0.8717
0.0	50.0	37500	1.6842	0.8717

Framework versions

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