smids_3x_beit_base_sgd_001_fold4

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3939
• Accuracy: 0.8483

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
0.8245	1.0	225	0.8340	0.6383
0.6387	2.0	450	0.6132	0.7483
0.532	3.0	675	0.5292	0.7867
0.4946	4.0	900	0.4935	0.8017
0.5105	5.0	1125	0.4602	0.8217
0.3964	6.0	1350	0.4420	0.8217
0.4068	7.0	1575	0.4284	0.83
0.4501	8.0	1800	0.4257	0.8217
0.3713	9.0	2025	0.4132	0.835
0.3427	10.0	2250	0.4081	0.8383
0.4054	11.0	2475	0.4089	0.8367
0.3818	12.0	2700	0.4017	0.84
0.3036	13.0	2925	0.4061	0.8317
0.2784	14.0	3150	0.3991	0.84
0.2822	15.0	3375	0.3953	0.8383
0.3106	16.0	3600	0.3913	0.8383
0.2716	17.0	3825	0.3985	0.8367
0.3166	18.0	4050	0.3943	0.8417
0.334	19.0	4275	0.3982	0.8333
0.2592	20.0	4500	0.3982	0.8383
0.2836	21.0	4725	0.3926	0.8367
0.2688	22.0	4950	0.3918	0.8417
0.2602	23.0	5175	0.3951	0.8417
0.2941	24.0	5400	0.3932	0.8417
0.254	25.0	5625	0.3963	0.8433
0.2248	26.0	5850	0.3967	0.8417
0.2349	27.0	6075	0.3902	0.8417
0.2318	28.0	6300	0.3960	0.8417
0.2339	29.0	6525	0.3900	0.8467
0.2256	30.0	6750	0.3940	0.8483
0.2306	31.0	6975	0.3948	0.84
0.1769	32.0	7200	0.3920	0.8433
0.2714	33.0	7425	0.3958	0.8483
0.2441	34.0	7650	0.3973	0.845
0.2336	35.0	7875	0.3946	0.8483
0.2411	36.0	8100	0.3957	0.8517
0.2513	37.0	8325	0.3968	0.845
0.2269	38.0	8550	0.3976	0.8467
0.2515	39.0	8775	0.3973	0.8517
0.2727	40.0	9000	0.3940	0.8467
0.2023	41.0	9225	0.3933	0.845
0.2359	42.0	9450	0.3953	0.85
0.2348	43.0	9675	0.3957	0.8483
0.2703	44.0	9900	0.3944	0.8517
0.2898	45.0	10125	0.3951	0.8483
0.2247	46.0	10350	0.3937	0.85
0.2326	47.0	10575	0.3934	0.85
0.2372	48.0	10800	0.3941	0.8483
0.2457	49.0	11025	0.3940	0.8483
0.2302	50.0	11250	0.3939	0.8483

Framework versions

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