beit-base-patch16-224-65-fold1

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: 0.3454
• Accuracy: 0.8732

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	0.7288	0.4648
No log	1.8462	6	0.6816	0.5634
No log	2.7692	9	0.6402	0.6479
0.699	4.0	13	0.6028	0.6620
0.699	4.9231	16	0.5877	0.6620
0.699	5.8462	19	0.5858	0.6620
0.618	6.7692	22	0.5099	0.7324
0.618	8.0	26	0.5926	0.6901
0.618	8.9231	29	0.4867	0.7746
0.5564	9.8462	32	0.4736	0.7746
0.5564	10.7692	35	0.4780	0.7606
0.5564	12.0	39	0.4812	0.7606
0.5017	12.9231	42	0.4834	0.7887
0.5017	13.8462	45	0.4662	0.8451
0.5017	14.7692	48	0.4560	0.8028
0.4603	16.0	52	0.4046	0.8169
0.4603	16.9231	55	0.3368	0.8451
0.4603	17.8462	58	0.3454	0.8732
0.4037	18.7692	61	0.6355	0.7324
0.4037	20.0	65	0.3624	0.8592
0.4037	20.9231	68	0.3748	0.8592
0.3696	21.8462	71	0.3799	0.8451
0.3696	22.7692	74	0.3886	0.8592
0.3696	24.0	78	0.3485	0.8451
0.3193	24.9231	81	0.3385	0.8732
0.3193	25.8462	84	0.3691	0.8592
0.3193	26.7692	87	0.3863	0.8732
0.308	28.0	91	0.3722	0.8732
0.308	28.9231	94	0.3481	0.8732
0.308	29.8462	97	0.3488	0.8732
0.2306	30.7692	100	0.4253	0.8451
0.2306	32.0	104	0.6244	0.7887
0.2306	32.9231	107	0.4688	0.7887
0.2431	33.8462	110	0.6080	0.7746
0.2431	34.7692	113	0.5795	0.7606
0.2431	36.0	117	0.5478	0.7887
0.2174	36.9231	120	0.5283	0.8169
0.2174	37.8462	123	0.5356	0.7887
0.2174	38.7692	126	0.5723	0.8169
0.1928	40.0	130	0.5489	0.8028
0.1928	40.9231	133	0.5277	0.7887
0.1928	41.8462	136	0.4725	0.8028
0.1928	42.7692	139	0.7954	0.7606
0.1919	44.0	143	0.5396	0.7887
0.1919	44.9231	146	0.6012	0.7746
0.1919	45.8462	149	0.6192	0.8028
0.1886	46.7692	152	0.6233	0.8028
0.1886	48.0	156	0.6465	0.8169
0.1886	48.9231	159	0.7676	0.8028
0.1661	49.8462	162	0.5266	0.8028
0.1661	50.7692	165	0.5127	0.8310
0.1661	52.0	169	0.5554	0.8169
0.1746	52.9231	172	0.6333	0.8310
0.1746	53.8462	175	0.6069	0.7887
0.1746	54.7692	178	0.6963	0.7887
0.1487	56.0	182	0.7242	0.8028
0.1487	56.9231	185	0.8501	0.7887
0.1487	57.8462	188	0.6207	0.7887
0.1785	58.7692	191	0.5805	0.7887
0.1785	60.0	195	0.5998	0.8169
0.1785	60.9231	198	0.5254	0.7887
0.1711	61.8462	201	0.5723	0.8028
0.1711	62.7692	204	0.7298	0.7887
0.1711	64.0	208	0.6647	0.7887
0.1637	64.9231	211	0.7011	0.8310
0.1637	65.8462	214	0.7031	0.8169
0.1637	66.7692	217	0.7163	0.8028
0.1618	68.0	221	0.6511	0.8169
0.1618	68.9231	224	0.6291	0.8310
0.1618	69.8462	227	0.6044	0.8451
0.153	70.7692	230	0.5888	0.8310
0.153	72.0	234	0.5881	0.8310
0.153	72.9231	237	0.5604	0.8169
0.1365	73.8462	240	0.6055	0.8310
0.1365	74.7692	243	0.6326	0.8028
0.1365	76.0	247	0.6686	0.7746
0.1231	76.9231	250	0.6955	0.7746
0.1231	77.8462	253	0.7302	0.7746
0.1231	78.7692	256	0.7928	0.7746
0.132	80.0	260	0.7247	0.7887
0.132	80.9231	263	0.7243	0.8028
0.132	81.8462	266	0.7361	0.8169
0.132	82.7692	269	0.7179	0.7746
0.1126	84.0	273	0.7054	0.7746
0.1126	84.9231	276	0.7192	0.7887
0.1126	85.8462	279	0.7358	0.7746
0.1141	86.7692	282	0.7575	0.7887
0.1141	88.0	286	0.7741	0.7887
0.1141	88.9231	289	0.7878	0.7887
0.1105	89.8462	292	0.7857	0.7887
0.1105	90.7692	295	0.7814	0.7887
0.1105	92.0	299	0.7785	0.7887
0.1006	92.3077	300	0.7782	0.7887

Framework versions

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