smids_10x_deit_small_adamax_001_fold2

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

• Loss: 1.0087
• Accuracy: 0.8985

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
0.3033	1.0	750	0.3519	0.8552
0.1821	2.0	1500	0.2712	0.9151
0.17	3.0	2250	0.2785	0.8985
0.2112	4.0	3000	0.3938	0.8735
0.1304	5.0	3750	0.4058	0.8719
0.1351	6.0	4500	0.3464	0.8835
0.0944	7.0	5250	0.4185	0.9018
0.08	8.0	6000	0.4321	0.9052
0.0621	9.0	6750	0.4755	0.8902
0.0267	10.0	7500	0.5448	0.8669
0.0347	11.0	8250	0.4694	0.8968
0.1206	12.0	9000	0.6302	0.8819
0.032	13.0	9750	0.6552	0.8819
0.0071	14.0	10500	0.5891	0.8869
0.0007	15.0	11250	0.6706	0.8935
0.0152	16.0	12000	0.6742	0.8852
0.0005	17.0	12750	0.6672	0.8952
0.0068	18.0	13500	0.7499	0.8885
0.0002	19.0	14250	0.7790	0.9018
0.0003	20.0	15000	0.7692	0.8835
0.0056	21.0	15750	0.8482	0.8752
0.0136	22.0	16500	0.8127	0.8835
0.0063	23.0	17250	0.6687	0.8952
0.0072	24.0	18000	0.8624	0.8869
0.0	25.0	18750	0.8382	0.8902
0.0001	26.0	19500	0.8780	0.8769
0.0107	27.0	20250	0.8313	0.8935
0.0	28.0	21000	0.9547	0.8869
0.0	29.0	21750	0.9878	0.8952
0.0	30.0	22500	0.8456	0.9035
0.0	31.0	23250	1.0397	0.8918
0.0	32.0	24000	0.9157	0.9018
0.0	33.0	24750	0.9451	0.9018
0.0	34.0	25500	0.9702	0.8985
0.0	35.0	26250	0.9002	0.9085
0.0	36.0	27000	0.9202	0.8985
0.004	37.0	27750	0.9200	0.8968
0.0	38.0	28500	0.9595	0.8968
0.0	39.0	29250	0.9656	0.9018
0.0	40.0	30000	0.9661	0.9018
0.0	41.0	30750	0.9523	0.9018
0.0	42.0	31500	0.9556	0.9002
0.0	43.0	32250	0.9641	0.8985
0.0	44.0	33000	0.9727	0.9002
0.0026	45.0	33750	0.9812	0.8985
0.0	46.0	34500	0.9958	0.8985
0.0	47.0	35250	0.9951	0.8985
0.0	48.0	36000	1.0005	0.8985
0.0	49.0	36750	1.0064	0.8985
0.0	50.0	37500	1.0087	0.8985

Framework versions

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