beit-base-patch16-224-75-fold3

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: 0.3326
• Accuracy: 0.9535

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	2	0.5223	0.8140
No log	2.0	4	0.5171	0.7907
No log	3.0	6	0.5562	0.7907
No log	4.0	8	0.5032	0.7907
0.5026	5.0	10	0.4636	0.7907
0.5026	6.0	12	0.4143	0.8372
0.5026	7.0	14	0.4338	0.8372
0.5026	8.0	16	0.4471	0.8372
0.5026	9.0	18	0.3882	0.8140
0.3578	10.0	20	0.3487	0.8372
0.3578	11.0	22	0.3975	0.8372
0.3578	12.0	24	0.3497	0.8372
0.3578	13.0	26	0.4257	0.8372
0.3578	14.0	28	0.2964	0.8837
0.322	15.0	30	0.4666	0.8372
0.322	16.0	32	0.3815	0.7907
0.322	17.0	34	0.2859	0.8837
0.322	18.0	36	0.5995	0.8140
0.322	19.0	38	0.4421	0.8372
0.2983	20.0	40	0.2857	0.9070
0.2983	21.0	42	0.3963	0.8605
0.2983	22.0	44	0.2775	0.9302
0.2983	23.0	46	0.4763	0.7442
0.2983	24.0	48	0.2789	0.8605
0.2545	25.0	50	0.3596	0.8372
0.2545	26.0	52	0.2447	0.9302
0.2545	27.0	54	0.4725	0.8140
0.2545	28.0	56	0.3389	0.8837
0.2545	29.0	58	0.3472	0.8140
0.1914	30.0	60	0.3326	0.9535
0.1914	31.0	62	0.2856	0.9302
0.1914	32.0	64	0.2857	0.8605
0.1914	33.0	66	0.2460	0.9302
0.1914	34.0	68	0.5097	0.8140
0.1782	35.0	70	0.2503	0.9535
0.1782	36.0	72	0.2338	0.9302
0.1782	37.0	74	0.4161	0.8837
0.1782	38.0	76	0.3522	0.9070
0.1782	39.0	78	0.2566	0.8372
0.1435	40.0	80	0.2658	0.8605
0.1435	41.0	82	0.2197	0.9302
0.1435	42.0	84	0.2200	0.8837
0.1435	43.0	86	0.2167	0.8837
0.1435	44.0	88	0.3905	0.9070
0.1131	45.0	90	0.5501	0.8837
0.1131	46.0	92	0.5251	0.9070
0.1131	47.0	94	0.4207	0.9070
0.1131	48.0	96	0.3181	0.8605
0.1131	49.0	98	0.4059	0.8140
0.1397	50.0	100	0.3061	0.8605
0.1397	51.0	102	0.2825	0.9302
0.1397	52.0	104	0.2886	0.9302
0.1397	53.0	106	0.3140	0.9070
0.1397	54.0	108	0.3363	0.9070
0.1048	55.0	110	0.3388	0.8837
0.1048	56.0	112	0.3391	0.9070
0.1048	57.0	114	0.3927	0.8605
0.1048	58.0	116	0.3799	0.8837
0.1048	59.0	118	0.5491	0.8140
0.1013	60.0	120	0.6218	0.8140
0.1013	61.0	122	0.4467	0.8372
0.1013	62.0	124	0.3842	0.9302
0.1013	63.0	126	0.3517	0.9302
0.1013	64.0	128	0.3083	0.9070
0.0987	65.0	130	0.3125	0.9535
0.0987	66.0	132	0.3968	0.9070
0.0987	67.0	134	0.3974	0.9070
0.0987	68.0	136	0.3304	0.9302
0.0987	69.0	138	0.3508	0.9070
0.0932	70.0	140	0.3760	0.9070
0.0932	71.0	142	0.3828	0.9535
0.0932	72.0	144	0.3291	0.9302
0.0932	73.0	146	0.3109	0.8837
0.0932	74.0	148	0.3078	0.9302
0.0917	75.0	150	0.4478	0.9070
0.0917	76.0	152	0.6594	0.8605
0.0917	77.0	154	0.6036	0.8605
0.0917	78.0	156	0.3920	0.9070
0.0917	79.0	158	0.3263	0.8605
0.0917	80.0	160	0.3551	0.8372
0.0917	81.0	162	0.3475	0.8372
0.0917	82.0	164	0.3338	0.8837
0.0917	83.0	166	0.3354	0.9070
0.0917	84.0	168	0.3611	0.9070
0.0901	85.0	170	0.3558	0.9070
0.0901	86.0	172	0.3412	0.8837
0.0901	87.0	174	0.3352	0.9302
0.0901	88.0	176	0.3627	0.9070
0.0901	89.0	178	0.3637	0.9070
0.0764	90.0	180	0.3369	0.9070
0.0764	91.0	182	0.3246	0.9302
0.0764	92.0	184	0.3208	0.9302
0.0764	93.0	186	0.3229	0.9302
0.0764	94.0	188	0.3326	0.9302
0.0664	95.0	190	0.3340	0.9302
0.0664	96.0	192	0.3296	0.9302
0.0664	97.0	194	0.3268	0.9302
0.0664	98.0	196	0.3260	0.9535
0.0664	99.0	198	0.3288	0.9535
0.057	100.0	200	0.3298	0.9535

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1