beit-base-patch16-224-65-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: 0.5711
• Accuracy: 0.8592

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	0.9231	3	0.8549	0.5211
No log	1.8462	6	0.6976	0.5634
No log	2.7692	9	0.6809	0.5634
0.7778	4.0	13	0.6459	0.6056
0.7778	4.9231	16	0.6353	0.6338
0.7778	5.8462	19	0.6141	0.6197
0.6542	6.7692	22	0.6003	0.6056
0.6542	8.0	26	0.6168	0.6761
0.6542	8.9231	29	0.5781	0.6901
0.5817	9.8462	32	0.5710	0.7324
0.5817	10.7692	35	0.5345	0.7465
0.5817	12.0	39	0.6058	0.6479
0.513	12.9231	42	0.6433	0.7042
0.513	13.8462	45	0.5830	0.7042
0.513	14.7692	48	0.6167	0.7042
0.4756	16.0	52	0.7304	0.6338
0.4756	16.9231	55	0.5485	0.7606
0.4756	17.8462	58	0.5166	0.7606
0.4123	18.7692	61	0.6267	0.7746
0.4123	20.0	65	0.4253	0.8169
0.4123	20.9231	68	0.4698	0.7746
0.3745	21.8462	71	0.5312	0.7887
0.3745	22.7692	74	0.5158	0.7465
0.3745	24.0	78	0.5969	0.8028
0.3751	24.9231	81	0.5419	0.7606
0.3751	25.8462	84	0.4630	0.8028
0.3751	26.7692	87	0.5367	0.8028
0.3079	28.0	91	0.5220	0.8310
0.3079	28.9231	94	0.5342	0.7887
0.3079	29.8462	97	0.5711	0.8592
0.2831	30.7692	100	0.5757	0.7606
0.2831	32.0	104	0.5200	0.7465
0.2831	32.9231	107	0.4496	0.8451
0.292	33.8462	110	0.6480	0.8169
0.292	34.7692	113	0.6956	0.7465
0.292	36.0	117	0.5629	0.8169
0.2712	36.9231	120	0.7614	0.6901
0.2712	37.8462	123	0.5625	0.8028
0.2712	38.7692	126	0.5711	0.7746
0.2447	40.0	130	0.5476	0.7746
0.2447	40.9231	133	0.5354	0.8028
0.2447	41.8462	136	0.5217	0.8169
0.2447	42.7692	139	0.5767	0.8028
0.185	44.0	143	0.5606	0.8169
0.185	44.9231	146	0.6719	0.7887
0.185	45.8462	149	0.6074	0.7887
0.1921	46.7692	152	0.6351	0.7746
0.1921	48.0	156	0.5916	0.7746
0.1921	48.9231	159	0.6103	0.7887
0.1844	49.8462	162	0.5758	0.7887
0.1844	50.7692	165	0.5497	0.8169
0.1844	52.0	169	0.5377	0.8310
0.17	52.9231	172	0.6279	0.8169
0.17	53.8462	175	0.5826	0.7887
0.17	54.7692	178	0.7173	0.7746
0.1724	56.0	182	0.5340	0.8451
0.1724	56.9231	185	0.5528	0.8592
0.1724	57.8462	188	0.6547	0.7887
0.1734	58.7692	191	0.5986	0.8310
0.1734	60.0	195	0.6057	0.8028
0.1734	60.9231	198	0.7183	0.8028
0.1582	61.8462	201	0.5912	0.8169
0.1582	62.7692	204	0.6002	0.8028
0.1582	64.0	208	0.7886	0.7606
0.1372	64.9231	211	0.7019	0.7887
0.1372	65.8462	214	0.6460	0.8169
0.1372	66.7692	217	0.6935	0.8028
0.153	68.0	221	0.8108	0.7746
0.153	68.9231	224	0.7539	0.7887
0.153	69.8462	227	0.7090	0.7746
0.1512	70.7692	230	0.7147	0.7887
0.1512	72.0	234	0.8680	0.8028
0.1512	72.9231	237	0.8785	0.7887
0.1381	73.8462	240	0.7413	0.7887
0.1381	74.7692	243	0.7255	0.8169
0.1381	76.0	247	0.7124	0.7887
0.1432	76.9231	250	0.7343	0.8028
0.1432	77.8462	253	0.7404	0.8028
0.1432	78.7692	256	0.6941	0.7887
0.1135	80.0	260	0.6721	0.8310
0.1135	80.9231	263	0.6692	0.8310
0.1135	81.8462	266	0.6880	0.8028
0.1135	82.7692	269	0.6857	0.8028
0.1182	84.0	273	0.6850	0.7887
0.1182	84.9231	276	0.6816	0.7887
0.1182	85.8462	279	0.7048	0.7746
0.1019	86.7692	282	0.7804	0.7746
0.1019	88.0	286	0.8013	0.7746
0.1019	88.9231	289	0.7506	0.7606
0.1163	89.8462	292	0.7047	0.7746
0.1163	90.7692	295	0.6763	0.8028
0.1163	92.0	299	0.6606	0.8028
0.1258	92.3077	300	0.6592	0.8028

Framework versions

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