smids_10x_deit_base_adamax_0001_fold2

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

• Loss: 1.0292
• 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: 0.0001
• 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
0.1628	1.0	750	0.2778	0.8885
0.0865	2.0	1500	0.3575	0.9002
0.0255	3.0	2250	0.5656	0.8885
0.0266	4.0	3000	0.5733	0.9002
0.0169	5.0	3750	0.7037	0.9002
0.0024	6.0	4500	0.6843	0.8918
0.0124	7.0	5250	0.7219	0.8985
0.0002	8.0	6000	0.8057	0.8935
0.002	9.0	6750	0.7964	0.8985
0.0	10.0	7500	0.7467	0.9002
0.0004	11.0	8250	0.6482	0.9002
0.0	12.0	9000	0.8380	0.8968
0.0	13.0	9750	0.8814	0.8902
0.0	14.0	10500	0.8400	0.8985
0.0	15.0	11250	0.7646	0.9068
0.0003	16.0	12000	0.8183	0.9068
0.0	17.0	12750	0.8397	0.8985
0.0	18.0	13500	0.9452	0.9052
0.0	19.0	14250	0.8230	0.9085
0.0	20.0	15000	0.8880	0.9085
0.0	21.0	15750	0.9221	0.9035
0.004	22.0	16500	0.9029	0.8968
0.0026	23.0	17250	0.9288	0.9002
0.0	24.0	18000	0.9054	0.9002
0.0	25.0	18750	0.9636	0.8952
0.0	26.0	19500	0.9715	0.8935
0.0035	27.0	20250	0.9914	0.9002
0.0	28.0	21000	1.0064	0.9002
0.0	29.0	21750	0.9401	0.8985
0.0	30.0	22500	0.9859	0.9002
0.0	31.0	23250	1.0344	0.8952
0.0	32.0	24000	1.0163	0.8935
0.0	33.0	24750	0.9803	0.9018
0.0	34.0	25500	1.0043	0.8918
0.0	35.0	26250	1.0148	0.8968
0.0	36.0	27000	1.0177	0.9035
0.0028	37.0	27750	1.0101	0.8985
0.0	38.0	28500	0.9955	0.9018
0.0	39.0	29250	1.0034	0.8985
0.0	40.0	30000	1.0043	0.8985
0.0	41.0	30750	0.9850	0.9002
0.0	42.0	31500	0.9989	0.8985
0.0	43.0	32250	1.0060	0.8985
0.0	44.0	33000	1.0153	0.8985
0.0024	45.0	33750	1.0191	0.8985
0.0	46.0	34500	1.0252	0.8985
0.0	47.0	35250	1.0250	0.8985
0.0	48.0	36000	1.0256	0.8985
0.0	49.0	36750	1.0285	0.8985
0.0	50.0	37500	1.0292	0.8985

Framework versions

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