smids_10x_deit_tiny_rms_001_fold5

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: 1.1358
• Accuracy: 0.77

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
0.842	1.0	750	0.7944	0.635
0.7921	2.0	1500	0.7461	0.68
0.7457	3.0	2250	0.7489	0.6567
0.7198	4.0	3000	0.6696	0.7017
0.7308	5.0	3750	0.6733	0.7117
0.6476	6.0	4500	0.6584	0.7183
0.6495	7.0	5250	0.6399	0.72
0.6634	8.0	6000	0.6560	0.6933
0.7106	9.0	6750	0.6143	0.7217
0.6252	10.0	7500	0.6122	0.7117
0.622	11.0	8250	0.5967	0.7217
0.5747	12.0	9000	0.6620	0.6833
0.5895	13.0	9750	0.5480	0.7533
0.5822	14.0	10500	0.5552	0.7517
0.5153	15.0	11250	0.5659	0.7583
0.6055	16.0	12000	0.6107	0.7233
0.575	17.0	12750	0.5677	0.7617
0.5736	18.0	13500	0.5602	0.7667
0.5782	19.0	14250	0.5634	0.76
0.6129	20.0	15000	0.5635	0.745
0.5336	21.0	15750	0.5596	0.755
0.506	22.0	16500	0.5757	0.76
0.524	23.0	17250	0.5491	0.7817
0.4616	24.0	18000	0.5444	0.775
0.5681	25.0	18750	0.5513	0.775
0.5138	26.0	19500	0.5393	0.77
0.3668	27.0	20250	0.5531	0.7683
0.4576	28.0	21000	0.5461	0.7833
0.4869	29.0	21750	0.5490	0.7817
0.4448	30.0	22500	0.5673	0.7817
0.4739	31.0	23250	0.5856	0.7717
0.3935	32.0	24000	0.5695	0.7983
0.4839	33.0	24750	0.5444	0.7983
0.3678	34.0	25500	0.5927	0.77
0.3843	35.0	26250	0.5986	0.7833
0.4018	36.0	27000	0.6231	0.7783
0.3249	37.0	27750	0.6467	0.7483
0.3738	38.0	28500	0.7366	0.76
0.3927	39.0	29250	0.6338	0.7633
0.315	40.0	30000	0.6392	0.78
0.2962	41.0	30750	0.7177	0.775
0.2563	42.0	31500	0.7289	0.7717
0.2899	43.0	32250	0.7576	0.7733
0.2733	44.0	33000	0.7845	0.7717
0.2911	45.0	33750	0.8279	0.77
0.2308	46.0	34500	0.8639	0.7767
0.2511	47.0	35250	0.9705	0.7667
0.1763	48.0	36000	1.0471	0.7633
0.1753	49.0	36750	1.1025	0.775
0.1437	50.0	37500	1.1358	0.77

Framework versions

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