hushem_5x_deit_tiny_sgd_001_fold1

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: 1.3348
• Accuracy: 0.4222

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
1.4407	1.0	27	1.4410	0.2667
1.382	2.0	54	1.4092	0.2667
1.3657	3.0	81	1.3896	0.2889
1.3344	4.0	108	1.3798	0.3111
1.2537	5.0	135	1.3657	0.3111
1.2237	6.0	162	1.3512	0.3333
1.2007	7.0	189	1.3419	0.3333
1.1972	8.0	216	1.3343	0.3556
1.1733	9.0	243	1.3294	0.3556
1.1145	10.0	270	1.3216	0.3778
1.1155	11.0	297	1.3166	0.3778
1.036	12.0	324	1.3103	0.3778
1.0595	13.0	351	1.3110	0.3778
1.0374	14.0	378	1.3072	0.3556
1.0659	15.0	405	1.3094	0.3778
1.034	16.0	432	1.3075	0.3778
0.954	17.0	459	1.3079	0.3778
0.9531	18.0	486	1.3085	0.3778
0.957	19.0	513	1.3124	0.3778
0.8978	20.0	540	1.3062	0.3778
0.885	21.0	567	1.3090	0.3778
0.8819	22.0	594	1.3143	0.3778
0.8801	23.0	621	1.3162	0.3778
0.857	24.0	648	1.3096	0.3778
0.8479	25.0	675	1.3101	0.3778
0.8743	26.0	702	1.3127	0.4
0.8288	27.0	729	1.3204	0.4
0.8104	28.0	756	1.3212	0.4
0.8245	29.0	783	1.3255	0.4
0.8139	30.0	810	1.3165	0.4
0.796	31.0	837	1.3232	0.4
0.7919	32.0	864	1.3216	0.4
0.7796	33.0	891	1.3211	0.4
0.7571	34.0	918	1.3245	0.4222
0.7521	35.0	945	1.3258	0.4222
0.7479	36.0	972	1.3280	0.4222
0.7621	37.0	999	1.3305	0.4222
0.7058	38.0	1026	1.3305	0.4222
0.71	39.0	1053	1.3319	0.4222
0.7228	40.0	1080	1.3314	0.4222
0.7146	41.0	1107	1.3317	0.4222
0.7189	42.0	1134	1.3333	0.4222
0.7405	43.0	1161	1.3338	0.4222
0.6872	44.0	1188	1.3335	0.4222
0.6996	45.0	1215	1.3344	0.4222
0.6979	46.0	1242	1.3345	0.4222
0.7035	47.0	1269	1.3347	0.4222
0.703	48.0	1296	1.3347	0.4222
0.7245	49.0	1323	1.3348	0.4222
0.7019	50.0	1350	1.3348	0.4222

Framework versions

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