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: 0.2149
• Accuracy: 0.9323

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: 200
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	30	0.2601	0.9350
No log	2.0	60	0.2320	0.9364
No log	3.0	90	0.2134	0.9405
0.2387	4.0	120	0.2026	0.9405
0.2387	5.0	150	0.1922	0.9405
0.2387	6.0	180	0.1880	0.9432
0.1643	7.0	210	0.1821	0.9432
0.1643	8.0	240	0.1779	0.9432
0.1643	9.0	270	0.1762	0.9432
0.1335	10.0	300	0.1731	0.9459
0.1335	11.0	330	0.1698	0.9445
0.1335	12.0	360	0.1676	0.9445
0.1335	13.0	390	0.1664	0.9445
0.1141	14.0	420	0.1659	0.9459
0.1141	15.0	450	0.1638	0.9459
0.1141	16.0	480	0.1626	0.9472
0.1029	17.0	510	0.1617	0.9472
0.1029	18.0	540	0.1608	0.9472
0.1029	19.0	570	0.1599	0.9486
0.0943	20.0	600	0.1605	0.9486
0.0943	21.0	630	0.1587	0.9486
0.0943	22.0	660	0.1584	0.9459
0.0943	23.0	690	0.1580	0.9486
0.0883	24.0	720	0.1577	0.9459
0.0883	25.0	750	0.1577	0.9459
0.0883	26.0	780	0.1575	0.9472
0.0849	27.0	810	0.1572	0.9459
0.0849	28.0	840	0.1569	0.9459
0.0849	29.0	870	0.1569	0.9459
0.0834	30.0	900	0.1569	0.9459

Framework versions

• Transformers 5.4.0
• Pytorch 2.11.0+cu128
• Datasets 4.8.4
• Tokenizers 0.22.2