hushem_1x_beit_base_adamax_001_fold2

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: 1.3645
• Accuracy: 0.5556

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	1.4123	0.2444
1.8567	2.0	12	1.3969	0.2444
1.8567	3.0	18	1.3773	0.4
1.4001	4.0	24	1.3688	0.3778
1.3691	5.0	30	1.3640	0.2444
1.3691	6.0	36	1.2556	0.5111
1.3116	7.0	42	1.4009	0.2667
1.3116	8.0	48	1.2324	0.4222
1.1799	9.0	54	1.1289	0.5111
1.1098	10.0	60	1.5348	0.2667
1.1098	11.0	66	1.2341	0.4222
1.0933	12.0	72	1.3191	0.4667
1.0933	13.0	78	1.3567	0.4
0.986	14.0	84	1.1728	0.3778
0.9075	15.0	90	1.1993	0.5111
0.9075	16.0	96	1.1869	0.3556
0.8205	17.0	102	1.3241	0.5333
0.8205	18.0	108	1.2073	0.5333
0.9036	19.0	114	1.2788	0.4889
0.7712	20.0	120	1.2208	0.4667
0.7712	21.0	126	1.2263	0.5333
0.6949	22.0	132	1.1609	0.4889
0.6949	23.0	138	1.1919	0.4222
0.7053	24.0	144	1.2190	0.5111
0.6439	25.0	150	1.2569	0.5556
0.6439	26.0	156	1.3636	0.5333
0.6537	27.0	162	1.4293	0.5778
0.6537	28.0	168	1.2396	0.5111
0.6181	29.0	174	1.3037	0.5556
0.5097	30.0	180	1.3049	0.5778
0.5097	31.0	186	1.1406	0.5333
0.5782	32.0	192	1.2396	0.5333
0.5782	33.0	198	1.2877	0.5111
0.5897	34.0	204	1.3944	0.5778
0.4972	35.0	210	1.2439	0.5556
0.4972	36.0	216	1.2993	0.5556
0.4729	37.0	222	1.3034	0.5556
0.4729	38.0	228	1.3631	0.5556
0.3719	39.0	234	1.4220	0.5778
0.4329	40.0	240	1.3836	0.5111
0.4329	41.0	246	1.3661	0.5556
0.3819	42.0	252	1.3645	0.5556
0.3819	43.0	258	1.3645	0.5556
0.3664	44.0	264	1.3645	0.5556
0.4152	45.0	270	1.3645	0.5556
0.4152	46.0	276	1.3645	0.5556
0.3637	47.0	282	1.3645	0.5556
0.3637	48.0	288	1.3645	0.5556
0.394	49.0	294	1.3645	0.5556
0.3776	50.0	300	1.3645	0.5556

Framework versions

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