beit-base-patch16-224-65-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: 0.4937
• Accuracy: 0.9014

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	0.7742	0.4507
No log	1.8462	6	0.7185	0.4930
No log	2.7692	9	0.6625	0.5634
0.7338	4.0	13	0.6136	0.7183
0.7338	4.9231	16	0.5974	0.6479
0.7338	5.8462	19	0.5771	0.6338
0.6191	6.7692	22	0.5400	0.7042
0.6191	8.0	26	0.5127	0.7183
0.6191	8.9231	29	0.5341	0.7324
0.5723	9.8462	32	0.4877	0.7887
0.5723	10.7692	35	0.6659	0.6197
0.5723	12.0	39	0.5790	0.6761
0.5161	12.9231	42	0.5001	0.7606
0.5161	13.8462	45	0.4195	0.8310
0.5161	14.7692	48	0.4806	0.7746
0.4982	16.0	52	0.4013	0.8028
0.4982	16.9231	55	0.4189	0.8028
0.4982	17.8462	58	0.4018	0.8310
0.438	18.7692	61	0.5230	0.7183
0.438	20.0	65	0.4768	0.7465
0.438	20.9231	68	0.4428	0.7887
0.4641	21.8462	71	0.4122	0.8169
0.4641	22.7692	74	0.4537	0.7746
0.4641	24.0	78	0.3838	0.8310
0.308	24.9231	81	0.4586	0.8028
0.308	25.8462	84	0.5623	0.8028
0.308	26.7692	87	0.4050	0.8310
0.2766	28.0	91	0.3860	0.8169
0.2766	28.9231	94	0.4062	0.8169
0.2766	29.8462	97	0.6191	0.8169
0.288	30.7692	100	0.6076	0.7746
0.288	32.0	104	0.5300	0.8169
0.288	32.9231	107	0.6178	0.7606
0.2676	33.8462	110	0.4465	0.8451
0.2676	34.7692	113	0.5893	0.7606
0.2676	36.0	117	0.4782	0.8169
0.2306	36.9231	120	0.4946	0.8310
0.2306	37.8462	123	0.4534	0.8451
0.2306	38.7692	126	0.4603	0.8451
0.2095	40.0	130	0.5839	0.8028
0.2095	40.9231	133	0.4536	0.8310
0.2095	41.8462	136	0.4617	0.8592
0.2095	42.7692	139	0.4531	0.8592
0.2171	44.0	143	0.4325	0.8732
0.2171	44.9231	146	0.4732	0.8592
0.2171	45.8462	149	0.4779	0.8592
0.1686	46.7692	152	0.4841	0.8451
0.1686	48.0	156	0.5690	0.8310
0.1686	48.9231	159	0.5477	0.8451
0.1644	49.8462	162	0.5844	0.8310
0.1644	50.7692	165	0.5818	0.8310
0.1644	52.0	169	0.4674	0.8451
0.1915	52.9231	172	0.5320	0.8732
0.1915	53.8462	175	0.4933	0.8451
0.1915	54.7692	178	0.5090	0.8592
0.1561	56.0	182	0.4864	0.8451
0.1561	56.9231	185	0.4652	0.8732
0.1561	57.8462	188	0.5113	0.8592
0.1298	58.7692	191	0.4803	0.8732
0.1298	60.0	195	0.4794	0.8451
0.1298	60.9231	198	0.4743	0.8451
0.1467	61.8462	201	0.4739	0.8592
0.1467	62.7692	204	0.5211	0.8451
0.1467	64.0	208	0.5315	0.8592
0.1363	64.9231	211	0.5182	0.8592
0.1363	65.8462	214	0.5160	0.8451
0.1363	66.7692	217	0.6170	0.8169
0.154	68.0	221	0.4857	0.8592
0.154	68.9231	224	0.4763	0.8592
0.154	69.8462	227	0.4937	0.9014
0.141	70.7692	230	0.5038	0.8873
0.141	72.0	234	0.5026	0.8592
0.141	72.9231	237	0.5019	0.8592
0.1166	73.8462	240	0.5028	0.8592
0.1166	74.7692	243	0.5226	0.8592
0.1166	76.0	247	0.5295	0.8732
0.117	76.9231	250	0.5073	0.8732
0.117	77.8462	253	0.5081	0.8732
0.117	78.7692	256	0.5036	0.8592
0.1037	80.0	260	0.5038	0.8451
0.1037	80.9231	263	0.5072	0.8451
0.1037	81.8462	266	0.5081	0.8451
0.1037	82.7692	269	0.5062	0.8310
0.1085	84.0	273	0.5144	0.8451
0.1085	84.9231	276	0.5208	0.8592
0.1085	85.8462	279	0.5248	0.8592
0.0939	86.7692	282	0.5301	0.8592
0.0939	88.0	286	0.5357	0.8451
0.0939	88.9231	289	0.5398	0.8451
0.0962	89.8462	292	0.5434	0.8451
0.0962	90.7692	295	0.5455	0.8451
0.0962	92.0	299	0.5448	0.8451
0.1131	92.3077	300	0.5446	0.8451

Framework versions

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