hushem_5x_deit_small_adamax_001_fold4

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

• Loss: 2.6296
• Accuracy: 0.7143

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
1.4114	1.0	28	1.1807	0.5238
1.2764	2.0	56	0.9991	0.5
0.9708	3.0	84	0.8093	0.5952
0.8975	4.0	112	0.7127	0.7143
0.8745	5.0	140	0.6175	0.7857
0.8263	6.0	168	0.5662	0.7857
0.7023	7.0	196	0.9422	0.4762
0.739	8.0	224	1.0785	0.6667
0.6219	9.0	252	0.7234	0.7381
0.5557	10.0	280	0.9010	0.6429
0.5536	11.0	308	0.8428	0.6905
0.4341	12.0	336	1.4219	0.5952
0.4008	13.0	364	1.0252	0.6190
0.3661	14.0	392	0.9782	0.6667
0.2255	15.0	420	1.0569	0.6905
0.1906	16.0	448	1.2300	0.7381
0.1975	17.0	476	1.3396	0.6905
0.1887	18.0	504	1.2504	0.6905
0.1207	19.0	532	1.3305	0.6905
0.0442	20.0	560	1.3516	0.7143
0.0689	21.0	588	1.7387	0.7143
0.0525	22.0	616	1.7134	0.6667
0.0339	23.0	644	1.8502	0.7381
0.0226	24.0	672	2.0846	0.7143
0.0026	25.0	700	2.2517	0.6905
0.0003	26.0	728	2.4690	0.6667
0.0231	27.0	756	2.7345	0.7143
0.0001	28.0	784	2.7167	0.7381
0.0001	29.0	812	2.5959	0.7143
0.0033	30.0	840	2.4481	0.7143
0.0096	31.0	868	2.5344	0.6905
0.0001	32.0	896	2.5507	0.7143
0.0	33.0	924	2.5703	0.7143
0.0	34.0	952	2.5803	0.7143
0.0	35.0	980	2.5893	0.7143
0.0	36.0	1008	2.5965	0.7143
0.0	37.0	1036	2.6022	0.7143
0.0	38.0	1064	2.6075	0.7143
0.0	39.0	1092	2.6116	0.7143
0.0	40.0	1120	2.6157	0.7143
0.0	41.0	1148	2.6188	0.7143
0.0	42.0	1176	2.6214	0.7143
0.0	43.0	1204	2.6241	0.7143
0.0	44.0	1232	2.6260	0.7143
0.0	45.0	1260	2.6276	0.7143
0.0	46.0	1288	2.6284	0.7143
0.0	47.0	1316	2.6292	0.7143
0.0	48.0	1344	2.6296	0.7143
0.0	49.0	1372	2.6296	0.7143
0.0	50.0	1400	2.6296	0.7143

Framework versions

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