hushem_5x_deit_small_sgd_001_fold2

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: 1.3629
• Accuracy: 0.4444

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
1.4659	1.0	27	1.4409	0.2889
1.3866	2.0	54	1.4026	0.3556
1.3486	3.0	81	1.3816	0.3333
1.3477	4.0	108	1.3676	0.3111
1.2816	5.0	135	1.3557	0.3333
1.2558	6.0	162	1.3444	0.3556
1.2259	7.0	189	1.3343	0.3556
1.2042	8.0	216	1.3245	0.3556
1.1683	9.0	243	1.3158	0.4
1.1515	10.0	270	1.3086	0.4222
1.1156	11.0	297	1.3037	0.4222
1.1061	12.0	324	1.2999	0.4444
1.0903	13.0	351	1.3002	0.4444
1.0661	14.0	378	1.3028	0.4444
1.0598	15.0	405	1.3085	0.4444
1.0378	16.0	432	1.3130	0.4444
1.0191	17.0	459	1.3179	0.4444
0.9884	18.0	486	1.3238	0.4444
0.9629	19.0	513	1.3282	0.4444
0.9575	20.0	540	1.3319	0.4222
0.9397	21.0	567	1.3353	0.4222
0.9296	22.0	594	1.3380	0.4222
0.9149	23.0	621	1.3408	0.4222
0.9023	24.0	648	1.3446	0.4222
0.8747	25.0	675	1.3454	0.4667
0.9184	26.0	702	1.3472	0.4444
0.8454	27.0	729	1.3479	0.4444
0.8505	28.0	756	1.3510	0.4444
0.8567	29.0	783	1.3517	0.4444
0.8854	30.0	810	1.3544	0.4667
0.834	31.0	837	1.3546	0.4444
0.8438	32.0	864	1.3560	0.4444
0.8236	33.0	891	1.3564	0.4444
0.8208	34.0	918	1.3570	0.4444
0.8066	35.0	945	1.3589	0.4444
0.8073	36.0	972	1.3591	0.4444
0.8089	37.0	999	1.3595	0.4444
0.777	38.0	1026	1.3599	0.4444
0.7828	39.0	1053	1.3610	0.4444
0.787	40.0	1080	1.3609	0.4444
0.8016	41.0	1107	1.3612	0.4444
0.7822	42.0	1134	1.3619	0.4444
0.8105	43.0	1161	1.3621	0.4444
0.7646	44.0	1188	1.3622	0.4444
0.7928	45.0	1215	1.3624	0.4444
0.7714	46.0	1242	1.3625	0.4444
0.7741	47.0	1269	1.3627	0.4444
0.7688	48.0	1296	1.3629	0.4444
0.834	49.0	1323	1.3629	0.4444
0.7751	50.0	1350	1.3629	0.4444

Framework versions

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