smids_3x_deit_tiny_rms_001_fold4

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: 2.2714
• Accuracy: 0.7733

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
0.8968	1.0	225	0.7639	0.6083
0.8626	2.0	450	0.7483	0.6217
0.7883	3.0	675	0.8572	0.525
0.7849	4.0	900	1.5683	0.415
0.8414	5.0	1125	0.7843	0.605
0.7458	6.0	1350	0.6775	0.6783
0.7372	7.0	1575	0.7834	0.63
0.7187	8.0	1800	0.6584	0.6983
0.6863	9.0	2025	0.6503	0.6817
0.6631	10.0	2250	0.6660	0.6967
0.6649	11.0	2475	0.6379	0.725
0.6942	12.0	2700	0.6307	0.7233
0.6966	13.0	2925	0.6415	0.735
0.5939	14.0	3150	0.6031	0.7383
0.6282	15.0	3375	0.6088	0.7217
0.602	16.0	3600	0.6243	0.7367
0.5204	17.0	3825	0.6588	0.7317
0.5969	18.0	4050	0.5604	0.7533
0.5806	19.0	4275	0.5450	0.7517
0.5391	20.0	4500	0.5377	0.7517
0.5038	21.0	4725	0.5743	0.7633
0.5713	22.0	4950	0.5280	0.7683
0.5554	23.0	5175	0.6486	0.74
0.5176	24.0	5400	0.5324	0.775
0.495	25.0	5625	0.5436	0.7867
0.5361	26.0	5850	0.5289	0.7717
0.4994	27.0	6075	0.5964	0.765
0.4408	28.0	6300	0.5977	0.755
0.4143	29.0	6525	0.5955	0.77
0.4019	30.0	6750	0.5477	0.79
0.3202	31.0	6975	0.6202	0.7733
0.2909	32.0	7200	0.6279	0.7667
0.3341	33.0	7425	0.6209	0.7933
0.3201	34.0	7650	0.6151	0.7883
0.3077	35.0	7875	0.6910	0.79
0.2136	36.0	8100	0.8074	0.7567
0.1784	37.0	8325	0.7942	0.785
0.2068	38.0	8550	0.8891	0.78
0.1962	39.0	8775	0.8894	0.775
0.1531	40.0	9000	1.1161	0.7783
0.0708	41.0	9225	1.1305	0.7867
0.0729	42.0	9450	1.2984	0.755
0.0781	43.0	9675	1.3627	0.7867
0.0522	44.0	9900	1.6222	0.7683
0.0281	45.0	10125	1.8871	0.7683
0.0088	46.0	10350	2.0271	0.7767
0.0017	47.0	10575	2.0484	0.7717
0.002	48.0	10800	2.2088	0.7767
0.0003	49.0	11025	2.2142	0.7717
0.0001	50.0	11250	2.2714	0.7733

Framework versions

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