beit-base-patch16-224-75-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: 0.2664
• 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	0.6862	0.5116
No log	2.0	4	0.5913	0.7209
No log	3.0	6	0.7204	0.6977
No log	4.0	8	0.5995	0.6977
0.6162	5.0	10	0.4235	0.8140
0.6162	6.0	12	0.3975	0.8140
0.6162	7.0	14	0.6029	0.7674
0.6162	8.0	16	0.4670	0.8140
0.6162	9.0	18	0.3448	0.8372
0.4312	10.0	20	0.4464	0.8372
0.4312	11.0	22	0.3396	0.8605
0.4312	12.0	24	0.4007	0.8372
0.4312	13.0	26	0.3398	0.8140
0.4312	14.0	28	0.4276	0.8605
0.3453	15.0	30	0.4336	0.8605
0.3453	16.0	32	0.3777	0.8140
0.3453	17.0	34	0.5910	0.8140
0.3453	18.0	36	0.6095	0.8140
0.3453	19.0	38	0.3570	0.8140
0.3288	20.0	40	0.5202	0.8140
0.3288	21.0	42	0.5604	0.8140
0.3288	22.0	44	0.2949	0.8372
0.3288	23.0	46	0.3442	0.8837
0.3288	24.0	48	0.2820	0.8372
0.2571	25.0	50	0.3240	0.8605
0.2571	26.0	52	0.2909	0.8837
0.2571	27.0	54	0.2429	0.8837
0.2571	28.0	56	0.2280	0.9302
0.2571	29.0	58	0.3984	0.8605
0.2012	30.0	60	0.2905	0.8605
0.2012	31.0	62	0.2509	0.9070
0.2012	32.0	64	0.2888	0.8605
0.2012	33.0	66	0.2689	0.8605
0.2012	34.0	68	0.2417	0.8837
0.1814	35.0	70	0.2418	0.9070
0.1814	36.0	72	0.2491	0.9070
0.1814	37.0	74	0.2998	0.9070
0.1814	38.0	76	0.2744	0.9302
0.1814	39.0	78	0.2664	0.9535
0.1555	40.0	80	0.2160	0.9302
0.1555	41.0	82	0.3875	0.9070
0.1555	42.0	84	0.4608	0.9070
0.1555	43.0	86	0.2978	0.9302
0.1555	44.0	88	0.4461	0.8837
0.1459	45.0	90	0.3603	0.9070
0.1459	46.0	92	0.2973	0.9302
0.1459	47.0	94	0.3385	0.8837
0.1459	48.0	96	0.3239	0.8837
0.1459	49.0	98	0.4315	0.8837
0.1372	50.0	100	0.3519	0.8837
0.1372	51.0	102	0.4148	0.8837
0.1372	52.0	104	0.4687	0.8837
0.1372	53.0	106	0.3287	0.8837
0.1372	54.0	108	0.3194	0.9070
0.1049	55.0	110	0.3703	0.8837
0.1049	56.0	112	0.3522	0.9070
0.1049	57.0	114	0.2572	0.9070
0.1049	58.0	116	0.2523	0.9070
0.1049	59.0	118	0.3136	0.9070
0.1143	60.0	120	0.3638	0.9070
0.1143	61.0	122	0.2916	0.9535
0.1143	62.0	124	0.2521	0.9302
0.1143	63.0	126	0.2735	0.9302
0.1143	64.0	128	0.3112	0.9302
0.0885	65.0	130	0.3246	0.9302
0.0885	66.0	132	0.3264	0.9070
0.0885	67.0	134	0.3351	0.9302
0.0885	68.0	136	0.3455	0.9302
0.0885	69.0	138	0.3579	0.9302
0.1064	70.0	140	0.3926	0.9302
0.1064	71.0	142	0.4370	0.9070
0.1064	72.0	144	0.4149	0.9302
0.1064	73.0	146	0.3315	0.9535
0.1064	74.0	148	0.2704	0.9302
0.1047	75.0	150	0.2600	0.9302
0.1047	76.0	152	0.3215	0.9535
0.1047	77.0	154	0.4110	0.9302
0.1047	78.0	156	0.4414	0.8837
0.1047	79.0	158	0.3589	0.9302
0.0937	80.0	160	0.3085	0.9535
0.0937	81.0	162	0.2889	0.9535
0.0937	82.0	164	0.2787	0.9535
0.0937	83.0	166	0.3251	0.9535
0.0937	84.0	168	0.4483	0.9070
0.0748	85.0	170	0.5490	0.8605
0.0748	86.0	172	0.5422	0.8605
0.0748	87.0	174	0.5282	0.8837
0.0748	88.0	176	0.5733	0.8605
0.0748	89.0	178	0.5978	0.8605
0.0834	90.0	180	0.5763	0.8605
0.0834	91.0	182	0.5270	0.8605
0.0834	92.0	184	0.4946	0.8837
0.0834	93.0	186	0.4881	0.9070
0.0834	94.0	188	0.5115	0.8605
0.1016	95.0	190	0.5445	0.8605
0.1016	96.0	192	0.5537	0.8605
0.1016	97.0	194	0.5451	0.8605
0.1016	98.0	196	0.5323	0.8605
0.1016	99.0	198	0.5190	0.8837
0.0657	100.0	200	0.5155	0.8837

Framework versions

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