hushem_5x_deit_base_rms_001_fold4

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: 1.5631
• Accuracy: 0.5714

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
2.0608	1.0	28	1.9020	0.2381
1.4315	2.0	56	1.5135	0.2619
1.432	3.0	84	1.3985	0.2143
1.6812	4.0	112	1.3928	0.2381
1.4086	5.0	140	1.5942	0.2619
1.4209	6.0	168	1.4007	0.2619
1.3906	7.0	196	1.3455	0.2857
1.3856	8.0	224	1.3904	0.3571
1.3883	9.0	252	1.5074	0.2619
1.3911	10.0	280	1.3463	0.3095
1.321	11.0	308	1.3274	0.4048
1.376	12.0	336	1.3115	0.5
1.321	13.0	364	1.3257	0.3333
1.2442	14.0	392	1.1218	0.5476
1.0941	15.0	420	1.2997	0.4286
1.1363	16.0	448	1.5465	0.4524
1.0638	17.0	476	0.9343	0.5952
1.0907	18.0	504	0.9902	0.5476
0.9963	19.0	532	0.8437	0.6190
1.0331	20.0	560	0.8320	0.6905
0.9883	21.0	588	1.0947	0.5714
0.9803	22.0	616	0.9089	0.5238
1.0008	23.0	644	0.8555	0.5952
0.9928	24.0	672	0.9537	0.5952
0.8708	25.0	700	0.9389	0.5476
0.8901	26.0	728	0.9330	0.5476
0.9805	27.0	756	1.0992	0.5476
0.9571	28.0	784	0.8585	0.5714
0.9235	29.0	812	0.9983	0.5476
0.906	30.0	840	0.8224	0.6667
0.9028	31.0	868	0.8413	0.5952
0.8341	32.0	896	0.8590	0.5714
0.8295	33.0	924	0.8367	0.6190
0.8624	34.0	952	0.8792	0.6429
0.813	35.0	980	0.8895	0.6190
0.7546	36.0	1008	0.9894	0.6190
0.7656	37.0	1036	0.9036	0.6429
0.7105	38.0	1064	0.9485	0.6190
0.6814	39.0	1092	1.0142	0.6667
0.6664	40.0	1120	0.9173	0.5238
0.6217	41.0	1148	1.2127	0.5476
0.6416	42.0	1176	0.9635	0.7143
0.5292	43.0	1204	0.9897	0.5952
0.4807	44.0	1232	1.1192	0.5714
0.3893	45.0	1260	1.3046	0.6190
0.4239	46.0	1288	1.5151	0.5952
0.3386	47.0	1316	1.5298	0.5714
0.3111	48.0	1344	1.5630	0.5714
0.2905	49.0	1372	1.5631	0.5714
0.3033	50.0	1400	1.5631	0.5714

Framework versions

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