smids_3x_deit_tiny_sgd_001_fold3

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: 0.3132
• Accuracy: 0.885

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
0.9334	1.0	225	0.9342	0.5617
0.6594	2.0	450	0.7339	0.7033
0.5574	3.0	675	0.6026	0.7783
0.5172	4.0	900	0.5194	0.82
0.484	5.0	1125	0.4765	0.83
0.4213	6.0	1350	0.4411	0.8317
0.3641	7.0	1575	0.4119	0.84
0.3918	8.0	1800	0.3965	0.8467
0.3897	9.0	2025	0.3767	0.855
0.3608	10.0	2250	0.3717	0.855
0.3065	11.0	2475	0.3577	0.8583
0.3443	12.0	2700	0.3519	0.8667
0.3224	13.0	2925	0.3447	0.8667
0.3208	14.0	3150	0.3383	0.8767
0.2793	15.0	3375	0.3349	0.8783
0.2956	16.0	3600	0.3330	0.875
0.3399	17.0	3825	0.3298	0.8783
0.2685	18.0	4050	0.3281	0.8733
0.2915	19.0	4275	0.3329	0.8767
0.2697	20.0	4500	0.3263	0.8767
0.2328	21.0	4725	0.3232	0.8767
0.2125	22.0	4950	0.3202	0.8783
0.2061	23.0	5175	0.3195	0.8817
0.2429	24.0	5400	0.3174	0.8833
0.2246	25.0	5625	0.3192	0.8867
0.2934	26.0	5850	0.3181	0.885
0.2259	27.0	6075	0.3160	0.885
0.2084	28.0	6300	0.3155	0.8833
0.2294	29.0	6525	0.3147	0.88
0.2001	30.0	6750	0.3145	0.8817
0.2211	31.0	6975	0.3141	0.8867
0.2219	32.0	7200	0.3133	0.8883
0.2508	33.0	7425	0.3138	0.8817
0.2086	34.0	7650	0.3129	0.8883
0.1887	35.0	7875	0.3125	0.8817
0.215	36.0	8100	0.3139	0.8867
0.1909	37.0	8325	0.3137	0.89
0.2136	38.0	8550	0.3150	0.8767
0.1884	39.0	8775	0.3120	0.8867
0.195	40.0	9000	0.3127	0.8833
0.178	41.0	9225	0.3134	0.885
0.1962	42.0	9450	0.3131	0.885
0.1627	43.0	9675	0.3128	0.885
0.1909	44.0	9900	0.3136	0.885
0.1981	45.0	10125	0.3133	0.885
0.1676	46.0	10350	0.3129	0.885
0.2286	47.0	10575	0.3130	0.885
0.1565	48.0	10800	0.3131	0.885
0.2098	49.0	11025	0.3132	0.885
0.1664	50.0	11250	0.3132	0.885

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2