smids_3x_deit_base_adamax_001_fold2

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: 1.0127
• Accuracy: 0.8869

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
0.4406	1.0	225	0.3562	0.8469
0.2499	2.0	450	0.3565	0.8486
0.2125	3.0	675	0.4018	0.8453
0.2418	4.0	900	0.3934	0.8569
0.1601	5.0	1125	0.3784	0.8586
0.1028	6.0	1350	0.4102	0.8669
0.1553	7.0	1575	0.4212	0.8602
0.0503	8.0	1800	0.4355	0.8835
0.1093	9.0	2025	0.4633	0.8752
0.0466	10.0	2250	0.4823	0.8769
0.0657	11.0	2475	0.5786	0.8686
0.0239	12.0	2700	0.4970	0.8835
0.0307	13.0	2925	0.5265	0.8686
0.0264	14.0	3150	0.5798	0.8935
0.0353	15.0	3375	0.6161	0.8835
0.0235	16.0	3600	0.6574	0.8852
0.0193	17.0	3825	0.6464	0.8869
0.0083	18.0	4050	0.5114	0.8935
0.0031	19.0	4275	0.6573	0.8869
0.0004	20.0	4500	0.6971	0.8918
0.023	21.0	4725	0.8443	0.8619
0.0243	22.0	4950	0.6663	0.8719
0.0379	23.0	5175	0.7440	0.8819
0.0041	24.0	5400	0.6714	0.8935
0.012	25.0	5625	0.8149	0.8802
0.0	26.0	5850	0.7898	0.8935
0.0001	27.0	6075	0.8193	0.8918
0.0	28.0	6300	0.7983	0.8852
0.0	29.0	6525	0.8430	0.8885
0.0044	30.0	6750	0.8519	0.8902
0.0027	31.0	6975	0.8733	0.8885
0.0	32.0	7200	0.8655	0.8935
0.0033	33.0	7425	0.8624	0.8852
0.0	34.0	7650	0.9256	0.8885
0.0	35.0	7875	0.9075	0.8852
0.0032	36.0	8100	0.9257	0.8869
0.0	37.0	8325	0.9450	0.8835
0.0	38.0	8550	0.9586	0.8819
0.0036	39.0	8775	0.9521	0.8852
0.0	40.0	9000	0.9863	0.8852
0.0044	41.0	9225	0.9719	0.8819
0.0	42.0	9450	0.9708	0.8819
0.0	43.0	9675	0.9922	0.8885
0.0	44.0	9900	0.9951	0.8869
0.0	45.0	10125	1.0055	0.8835
0.0	46.0	10350	1.0046	0.8885
0.0	47.0	10575	1.0098	0.8852
0.0	48.0	10800	1.0105	0.8885
0.0024	49.0	11025	1.0125	0.8869
0.0024	50.0	11250	1.0127	0.8869

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2