smids_3x_deit_tiny_rms_0001_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: 1.0441
• Accuracy: 0.89

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: 0.0001
• 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
0.384	1.0	225	0.3552	0.8583
0.2084	2.0	450	0.3480	0.8767
0.201	3.0	675	0.4132	0.8717
0.1086	4.0	900	0.3792	0.8883
0.1243	5.0	1125	0.4370	0.8733
0.1107	6.0	1350	0.4615	0.8783
0.0664	7.0	1575	0.6966	0.865
0.0782	8.0	1800	0.6855	0.8733
0.0473	9.0	2025	0.5753	0.875
0.0605	10.0	2250	0.7929	0.87
0.0782	11.0	2475	0.6012	0.8933
0.0086	12.0	2700	0.6879	0.8817
0.0115	13.0	2925	0.8156	0.86
0.0467	14.0	3150	1.0598	0.8467
0.1265	15.0	3375	0.7615	0.875
0.0039	16.0	3600	0.7484	0.8767
0.0138	17.0	3825	0.8169	0.87
0.0039	18.0	4050	0.8702	0.8783
0.0077	19.0	4275	0.8767	0.8867
0.0399	20.0	4500	0.8253	0.8817
0.0266	21.0	4725	1.0317	0.8567
0.0092	22.0	4950	1.0021	0.8683
0.0011	23.0	5175	0.9409	0.8867
0.0242	24.0	5400	0.9565	0.8733
0.0188	25.0	5625	0.8702	0.88
0.0079	26.0	5850	0.8620	0.8783
0.009	27.0	6075	0.8382	0.8883
0.0334	28.0	6300	0.8240	0.885
0.0091	29.0	6525	0.9309	0.88
0.0428	30.0	6750	0.8520	0.8817
0.0064	31.0	6975	0.9518	0.8833
0.0205	32.0	7200	0.8143	0.8983
0.0358	33.0	7425	1.0040	0.8867
0.0085	34.0	7650	0.9891	0.88
0.0	35.0	7875	0.9233	0.9
0.0	36.0	8100	0.9033	0.8917
0.0025	37.0	8325	0.9886	0.895
0.001	38.0	8550	1.1074	0.87
0.0	39.0	8775	1.0071	0.8883
0.0	40.0	9000	1.0033	0.8883
0.0	41.0	9225	1.0288	0.8867
0.0	42.0	9450	1.0506	0.8833
0.0	43.0	9675	1.0220	0.8883
0.0	44.0	9900	1.0225	0.885
0.0	45.0	10125	1.0227	0.8867
0.0	46.0	10350	1.0356	0.8867
0.0027	47.0	10575	1.0358	0.8917
0.0	48.0	10800	1.0430	0.89
0.0	49.0	11025	1.0451	0.89
0.0	50.0	11250	1.0441	0.89

Framework versions

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