smids_5x_deit_tiny_rms_0001_fold4

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

• Loss: 1.4357
• Accuracy: 0.8667

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
0.3287	1.0	375	0.3863	0.85
0.2455	2.0	750	0.3649	0.8717
0.1213	3.0	1125	0.4642	0.8583
0.1727	4.0	1500	0.5805	0.8617
0.1128	5.0	1875	0.6371	0.8483
0.0689	6.0	2250	0.6331	0.8683
0.0983	7.0	2625	0.6829	0.865
0.1105	8.0	3000	0.6645	0.8617
0.0716	9.0	3375	0.9136	0.8583
0.0639	10.0	3750	0.7869	0.8867
0.0325	11.0	4125	0.8744	0.8733
0.0627	12.0	4500	0.9757	0.8567
0.0409	13.0	4875	0.9654	0.8633
0.0848	14.0	5250	0.8074	0.8667
0.0374	15.0	5625	0.9236	0.8667
0.037	16.0	6000	1.0898	0.8617
0.0497	17.0	6375	1.1236	0.8583
0.0095	18.0	6750	1.0183	0.87
0.0289	19.0	7125	1.0208	0.8783
0.0255	20.0	7500	1.1375	0.8667
0.0016	21.0	7875	1.1251	0.8617
0.0005	22.0	8250	1.0252	0.8717
0.015	23.0	8625	1.1223	0.865
0.0375	24.0	9000	1.0372	0.8733
0.0379	25.0	9375	0.9869	0.8667
0.0001	26.0	9750	1.0331	0.8733
0.0134	27.0	10125	0.9754	0.885
0.0	28.0	10500	1.0742	0.8583
0.0001	29.0	10875	1.0378	0.88
0.0	30.0	11250	1.1203	0.875
0.0077	31.0	11625	1.1471	0.8783
0.0003	32.0	12000	1.1437	0.8783
0.0	33.0	12375	1.1521	0.875
0.0003	34.0	12750	1.2362	0.865
0.0	35.0	13125	1.2535	0.8567
0.0	36.0	13500	1.2428	0.865
0.0002	37.0	13875	1.3504	0.8583
0.0191	38.0	14250	1.2705	0.87
0.0	39.0	14625	1.3466	0.8667
0.0	40.0	15000	1.3575	0.8617
0.0	41.0	15375	1.3681	0.8667
0.0	42.0	15750	1.3681	0.87
0.0	43.0	16125	1.3799	0.865
0.0	44.0	16500	1.3559	0.8667
0.0	45.0	16875	1.3770	0.865
0.0	46.0	17250	1.4044	0.8667
0.0	47.0	17625	1.4188	0.8683
0.0	48.0	18000	1.4286	0.8667
0.0	49.0	18375	1.4343	0.8667
0.0	50.0	18750	1.4357	0.8667

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2