hushem_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: 1.1423
• Accuracy: 0.8571

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
1.5018	1.0	28	1.4435	0.2381
1.3915	2.0	56	1.7481	0.2619
0.9807	3.0	84	1.0427	0.5
0.6115	4.0	112	0.6071	0.6905
0.3048	5.0	140	0.8476	0.8095
0.2247	6.0	168	1.1744	0.7143
0.1107	7.0	196	0.4922	0.8810
0.0936	8.0	224	0.8411	0.8333
0.052	9.0	252	1.2457	0.7381
0.1043	10.0	280	0.5514	0.8810
0.0015	11.0	308	0.9844	0.8333
0.0204	12.0	336	0.4946	0.8333
0.0149	13.0	364	0.8042	0.8571
0.0016	14.0	392	1.2413	0.7857
0.0114	15.0	420	1.0274	0.8333
0.0352	16.0	448	0.9012	0.8095
0.0613	17.0	476	1.0834	0.7857
0.0277	18.0	504	0.9416	0.8571
0.0099	19.0	532	1.2649	0.7857
0.0251	20.0	560	0.9863	0.8571
0.0001	21.0	588	0.9424	0.8333
0.0001	22.0	616	0.9653	0.8571
0.0	23.0	644	0.9755	0.8571
0.0	24.0	672	0.9859	0.8571
0.0	25.0	700	0.9939	0.8571
0.0	26.0	728	1.0026	0.8571
0.0	27.0	756	1.0107	0.8571
0.0	28.0	784	1.0180	0.8571
0.0	29.0	812	1.0261	0.8571
0.0	30.0	840	1.0335	0.8571
0.0	31.0	868	1.0404	0.8571
0.0	32.0	896	1.0484	0.8571
0.0	33.0	924	1.0559	0.8571
0.0	34.0	952	1.0642	0.8571
0.0	35.0	980	1.0715	0.8571
0.0	36.0	1008	1.0789	0.8571
0.0	37.0	1036	1.0861	0.8571
0.0	38.0	1064	1.0922	0.8571
0.0	39.0	1092	1.0992	0.8571
0.0	40.0	1120	1.1058	0.8571
0.0	41.0	1148	1.1124	0.8571
0.0	42.0	1176	1.1189	0.8571
0.0	43.0	1204	1.1246	0.8571
0.0	44.0	1232	1.1302	0.8571
0.0	45.0	1260	1.1347	0.8571
0.0	46.0	1288	1.1383	0.8571
0.0	47.0	1316	1.1409	0.8571
0.0	48.0	1344	1.1422	0.8571
0.0	49.0	1372	1.1423	0.8571
0.0	50.0	1400	1.1423	0.8571

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0