BilalMuftuoglu/beit-base-patch16-224-55-fold3

beit-base-patch16-224-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: 0.5597
• Accuracy: 0.8354

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	0.8571	3	0.8417	0.4810
No log	2.0	7	0.6764	0.5823
0.71	2.8571	10	0.7272	0.5316
0.71	4.0	14	0.6122	0.6835
0.71	4.8571	17	0.6288	0.5949
0.6227	6.0	21	0.6550	0.5949
0.6227	6.8571	24	0.6240	0.6329
0.6227	8.0	28	0.5877	0.6709
0.5472	8.8571	31	0.7285	0.5823
0.5472	10.0	35	0.8305	0.5823
0.5472	10.8571	38	0.5102	0.7848
0.4766	12.0	42	0.5352	0.7215
0.4766	12.8571	45	0.5357	0.6962
0.4766	14.0	49	0.7418	0.6329
0.408	14.8571	52	0.6150	0.6835
0.408	16.0	56	0.4870	0.7975
0.408	16.8571	59	0.6427	0.6962
0.4078	18.0	63	0.4822	0.8101
0.4078	18.8571	66	0.4947	0.7975
0.3478	20.0	70	0.6847	0.7089
0.3478	20.8571	73	0.6154	0.7342
0.3478	22.0	77	0.5384	0.8101
0.3006	22.8571	80	0.5939	0.7595
0.3006	24.0	84	0.5214	0.7595
0.3006	24.8571	87	0.5452	0.7342
0.2977	26.0	91	0.6153	0.7215
0.2977	26.8571	94	0.4730	0.7975
0.2977	28.0	98	0.4861	0.7342
0.2768	28.8571	101	0.6705	0.7342
0.2768	30.0	105	0.6362	0.7848
0.2768	30.8571	108	0.6548	0.7848
0.2348	32.0	112	0.5100	0.7848
0.2348	32.8571	115	0.7156	0.7595
0.2348	34.0	119	0.4859	0.8228
0.2199	34.8571	122	0.8490	0.7342
0.2199	36.0	126	0.6095	0.7468
0.2199	36.8571	129	0.6427	0.7468
0.201	38.0	133	0.6283	0.7848
0.201	38.8571	136	0.8883	0.7595
0.1868	40.0	140	0.7146	0.7975
0.1868	40.8571	143	1.3800	0.6962
0.1868	42.0	147	0.5908	0.7848
0.2011	42.8571	150	0.6158	0.7722
0.2011	44.0	154	0.5477	0.7975
0.2011	44.8571	157	0.8354	0.7722
0.1807	46.0	161	0.7830	0.7848
0.1807	46.8571	164	0.6327	0.8228
0.1807	48.0	168	0.7858	0.7595
0.1579	48.8571	171	0.8322	0.7342
0.1579	50.0	175	0.7501	0.7848
0.1579	50.8571	178	0.8303	0.8228
0.2066	52.0	182	0.6831	0.7595
0.2066	52.8571	185	0.7837	0.8228
0.2066	54.0	189	0.5597	0.8354
0.1647	54.8571	192	0.5484	0.8354
0.1647	56.0	196	1.0047	0.7848
0.1647	56.8571	199	0.7815	0.8228
0.1404	58.0	203	0.6808	0.7975
0.1404	58.8571	206	1.0068	0.8101
0.1451	60.0	210	0.7698	0.8228
0.1451	60.8571	213	0.6495	0.8228
0.1451	62.0	217	0.7066	0.8354
0.1341	62.8571	220	0.6250	0.8354
0.1341	64.0	224	0.5573	0.7975
0.1341	64.8571	227	0.6051	0.8101
0.127	66.0	231	0.7576	0.8101
0.127	66.8571	234	0.8297	0.8101
0.127	68.0	238	1.0732	0.7975
0.1129	68.8571	241	1.0503	0.7975
0.1129	70.0	245	0.7520	0.8101
0.1129	70.8571	248	0.6825	0.8354
0.1205	72.0	252	0.7002	0.7975
0.1205	72.8571	255	0.7430	0.8101
0.1205	74.0	259	0.7610	0.7975
0.1199	74.8571	262	0.6854	0.8101
0.1199	76.0	266	0.6767	0.8354
0.1199	76.8571	269	0.6685	0.8354
0.1165	78.0	273	0.7134	0.7848
0.1165	78.8571	276	0.7344	0.7848
0.1213	80.0	280	0.7403	0.7722
0.1213	80.8571	283	0.7818	0.7848
0.1213	82.0	287	0.7620	0.7722
0.1024	82.8571	290	0.7539	0.7722
0.1024	84.0	294	0.7659	0.7722
0.1024	84.8571	297	0.7686	0.7848
0.1109	85.7143	300	0.7686	0.7848

Framework versions

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