vit-base-oxford-iiit-pets

This model is a fine-tuned version of google/vit-base-patch16-224 on the pcuenq/oxford-pets dataset. It achieves the following results on the evaluation set:

• Loss: 0.1798
• Accuracy: 0.9310

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: 0.0003
• train_batch_size: 512
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	12	2.6101	0.5223
No log	2.0	24	1.7190	0.8227
No log	3.0	36	1.0833	0.8890
No log	4.0	48	0.7011	0.9120
No log	5.0	60	0.5052	0.9242
No log	6.0	72	0.4097	0.9310
No log	7.0	84	0.3560	0.9350
No log	8.0	96	0.3237	0.9337
1.1364	9.0	108	0.3008	0.9378
1.1364	10.0	120	0.2833	0.9364
1.1364	11.0	132	0.2694	0.9391
1.1364	12.0	144	0.2586	0.9391
1.1364	13.0	156	0.2498	0.9418
1.1364	14.0	168	0.2423	0.9405
1.1364	15.0	180	0.2359	0.9405
1.1364	16.0	192	0.2303	0.9459
0.2326	17.0	204	0.2259	0.9405
0.2326	18.0	216	0.2222	0.9405
0.2326	19.0	228	0.2178	0.9432
0.2326	20.0	240	0.2146	0.9445
0.2326	21.0	252	0.2114	0.9432
0.2326	22.0	264	0.2087	0.9445
0.2326	23.0	276	0.2061	0.9432
0.2326	24.0	288	0.2040	0.9459
0.1651	25.0	300	0.2018	0.9459
0.1651	26.0	312	0.2000	0.9445
0.1651	27.0	324	0.1985	0.9459
0.1651	28.0	336	0.1968	0.9472
0.1651	29.0	348	0.1948	0.9459
0.1651	30.0	360	0.1939	0.9459
0.1651	31.0	372	0.1924	0.9459
0.1651	32.0	384	0.1915	0.9459
0.1651	33.0	396	0.1909	0.9459
0.134	34.0	408	0.1894	0.9472
0.134	35.0	420	0.1883	0.9459
0.134	36.0	432	0.1877	0.9472
0.134	37.0	444	0.1866	0.9486
0.134	38.0	456	0.1863	0.9472
0.134	39.0	468	0.1851	0.9486
0.134	40.0	480	0.1843	0.9472
0.134	41.0	492	0.1837	0.9472
0.1128	42.0	504	0.1831	0.9459
0.1128	43.0	516	0.1828	0.9472
0.1128	44.0	528	0.1822	0.9472
0.1128	45.0	540	0.1816	0.9472
0.1128	46.0	552	0.1808	0.9459
0.1128	47.0	564	0.1804	0.9459
0.1128	48.0	576	0.1802	0.9459
0.1128	49.0	588	0.1796	0.9459
0.0999	50.0	600	0.1793	0.9472
0.0999	51.0	612	0.1792	0.9486
0.0999	52.0	624	0.1787	0.9472
0.0999	53.0	636	0.1784	0.9472
0.0999	54.0	648	0.1780	0.9459
0.0999	55.0	660	0.1778	0.9445
0.0999	56.0	672	0.1772	0.9445
0.0999	57.0	684	0.1769	0.9472
0.0999	58.0	696	0.1768	0.9472
0.0894	59.0	708	0.1766	0.9472
0.0894	60.0	720	0.1763	0.9472
0.0894	61.0	732	0.1762	0.9486
0.0894	62.0	744	0.1760	0.9472
0.0894	63.0	756	0.1755	0.9459
0.0894	64.0	768	0.1752	0.9459
0.0894	65.0	780	0.1749	0.9459
0.0894	66.0	792	0.1749	0.9459
0.0828	67.0	804	0.1746	0.9472
0.0828	68.0	816	0.1745	0.9459
0.0828	69.0	828	0.1745	0.9459
0.0828	70.0	840	0.1744	0.9459
0.0828	71.0	852	0.1740	0.9459
0.0828	72.0	864	0.1741	0.9459
0.0828	73.0	876	0.1737	0.9459
0.0828	74.0	888	0.1739	0.9459
0.0778	75.0	900	0.1739	0.9459
0.0778	76.0	912	0.1737	0.9459
0.0778	77.0	924	0.1735	0.9459
0.0778	78.0	936	0.1733	0.9459
0.0778	79.0	948	0.1732	0.9459
0.0778	80.0	960	0.1732	0.9459
0.0778	81.0	972	0.1730	0.9459
0.0778	82.0	984	0.1730	0.9459
0.0778	83.0	996	0.1730	0.9459
0.0738	84.0	1008	0.1729	0.9459
0.0738	85.0	1020	0.1727	0.9459
0.0738	86.0	1032	0.1726	0.9459
0.0738	87.0	1044	0.1726	0.9459
0.0738	88.0	1056	0.1726	0.9459
0.0738	89.0	1068	0.1726	0.9459
0.0738	90.0	1080	0.1725	0.9459
0.0738	91.0	1092	0.1724	0.9459
0.0715	92.0	1104	0.1724	0.9459
0.0715	93.0	1116	0.1723	0.9459
0.0715	94.0	1128	0.1723	0.9459
0.0715	95.0	1140	0.1723	0.9459
0.0715	96.0	1152	0.1722	0.9459
0.0715	97.0	1164	0.1722	0.9459
0.0715	98.0	1176	0.1722	0.9459
0.0715	99.0	1188	0.1722	0.9459
0.0701	100.0	1200	0.1722	0.9459

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1
• Datasets 3.0.0
• Tokenizers 0.19.1