hushem_5x_deit_base_adamax_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: 1.8006
• Accuracy: 0.6905

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
1.4494	1.0	28	1.3719	0.2619
1.2522	2.0	56	0.9742	0.6667
1.0638	3.0	84	0.9282	0.4286
0.9326	4.0	112	0.9751	0.7381
0.9775	5.0	140	0.6128	0.8333
0.8386	6.0	168	0.6453	0.6905
0.7523	7.0	196	0.8760	0.5952
0.8483	8.0	224	0.6776	0.6905
0.7007	9.0	252	0.6406	0.7381
0.6736	10.0	280	1.1732	0.5714
0.6667	11.0	308	0.8999	0.7143
0.5535	12.0	336	0.7518	0.7143
0.5519	13.0	364	1.2198	0.6429
0.4746	14.0	392	1.2629	0.6190
0.4049	15.0	420	1.0670	0.7143
0.2485	16.0	448	1.3207	0.6667
0.2835	17.0	476	0.9080	0.7143
0.1908	18.0	504	0.9684	0.6905
0.1239	19.0	532	0.8600	0.8333
0.2177	20.0	560	1.2908	0.6667
0.0633	21.0	588	1.7014	0.7143
0.0847	22.0	616	1.3740	0.7857
0.1199	23.0	644	1.1620	0.8095
0.0618	24.0	672	1.7626	0.7857
0.0552	25.0	700	1.7596	0.7381
0.0166	26.0	728	1.4380	0.7143
0.0048	27.0	756	2.1450	0.6667
0.0064	28.0	784	1.7983	0.7381
0.0065	29.0	812	1.9453	0.6429
0.0052	30.0	840	1.5896	0.7619
0.0125	31.0	868	1.6540	0.7381
0.0008	32.0	896	1.7879	0.7619
0.0001	33.0	924	1.9506	0.7381
0.0002	34.0	952	1.7166	0.7143
0.0	35.0	980	1.7316	0.6905
0.0	36.0	1008	1.7446	0.6905
0.0	37.0	1036	1.7559	0.6905
0.0	38.0	1064	1.7638	0.6905
0.0	39.0	1092	1.7724	0.6905
0.0	40.0	1120	1.7784	0.6905
0.0	41.0	1148	1.7832	0.6905
0.0	42.0	1176	1.7877	0.6905
0.0	43.0	1204	1.7918	0.6905
0.0	44.0	1232	1.7950	0.6905
0.0	45.0	1260	1.7970	0.6905
0.0	46.0	1288	1.7988	0.6905
0.0	47.0	1316	1.8001	0.6905
0.0	48.0	1344	1.8006	0.6905
0.0	49.0	1372	1.8006	0.6905
0.0	50.0	1400	1.8006	0.6905

Framework versions

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