hushem_1x_beit_base_sgd_001_fold5

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: 1.3101
• Accuracy: 0.4390

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
No log	1.0	6	1.5856	0.2439
1.5446	2.0	12	1.5524	0.2683
1.5446	3.0	18	1.5246	0.3171
1.4921	4.0	24	1.5015	0.3171
1.4491	5.0	30	1.4859	0.3415
1.4491	6.0	36	1.4721	0.3415
1.4253	7.0	42	1.4615	0.3415
1.4253	8.0	48	1.4471	0.3659
1.3656	9.0	54	1.4347	0.3902
1.3889	10.0	60	1.4270	0.3902
1.3889	11.0	66	1.4192	0.4146
1.3303	12.0	72	1.4108	0.4146
1.3303	13.0	78	1.4040	0.4146
1.3227	14.0	84	1.3958	0.4146
1.3003	15.0	90	1.3889	0.4146
1.3003	16.0	96	1.3827	0.4146
1.3072	17.0	102	1.3788	0.3902
1.3072	18.0	108	1.3733	0.4146
1.2978	19.0	114	1.3664	0.4390
1.268	20.0	120	1.3623	0.4390
1.268	21.0	126	1.3569	0.4390
1.265	22.0	132	1.3511	0.4390
1.265	23.0	138	1.3470	0.4634
1.2559	24.0	144	1.3424	0.4390
1.2443	25.0	150	1.3395	0.4146
1.2443	26.0	156	1.3357	0.4390
1.2468	27.0	162	1.3318	0.4390
1.2468	28.0	168	1.3281	0.4390
1.2381	29.0	174	1.3262	0.4390
1.2466	30.0	180	1.3249	0.4146
1.2466	31.0	186	1.3215	0.4390
1.234	32.0	192	1.3185	0.4390
1.234	33.0	198	1.3170	0.4390
1.2144	34.0	204	1.3158	0.4390
1.2407	35.0	210	1.3143	0.4390
1.2407	36.0	216	1.3132	0.4390
1.2238	37.0	222	1.3125	0.4390
1.2238	38.0	228	1.3116	0.4390
1.221	39.0	234	1.3110	0.4390
1.1985	40.0	240	1.3104	0.4390
1.1985	41.0	246	1.3101	0.4390
1.2078	42.0	252	1.3101	0.4390
1.2078	43.0	258	1.3101	0.4390
1.1965	44.0	264	1.3101	0.4390
1.2151	45.0	270	1.3101	0.4390
1.2151	46.0	276	1.3101	0.4390
1.2187	47.0	282	1.3101	0.4390
1.2187	48.0	288	1.3101	0.4390
1.1908	49.0	294	1.3101	0.4390
1.1985	50.0	300	1.3101	0.4390

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0