results

This model is a fine-tuned version of google/siglip2-large-patch16-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 7.3621
• Model Preparation Time: 0.0027
• Age Mae: 3.9585
• Gender Acc: 0.9706
• Eth Acc: 0.8793
• Age Group Acc: 0.6400
• Gender Age Group Acc: 0.6197

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: 4e-05
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• distributed_type: tpu
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use OptimizerNames.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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Age Mae	Gender Acc	Eth Acc	Age Group Acc	Gender Age Group Acc
No log	1.0	170	8.1110	0.0027	5.2484	0.9552	0.8125	0.5699	0.5562
No log	2.0	340	6.7194	0.0027	4.2664	0.9672	0.8486	0.6114	0.5915
8.9853	3.0	510	7.1145	0.0027	4.6598	0.9664	0.8627	0.6027	0.5881
8.9853	4.0	680	6.8165	0.0027	4.4121	0.9697	0.8590	0.6101	0.5844
8.9853	5.0	850	6.4416	0.0027	4.1299	0.9685	0.8714	0.6267	0.6006
6.2831	6.0	1020	6.3622	0.0027	4.1352	0.9689	0.8764	0.6491	0.6242
6.2831	7.0	1190	6.7311	0.0027	4.4327	0.9706	0.8735	0.6217	0.6043
6.2831	8.0	1360	6.3400	0.0027	4.1008	0.9722	0.8727	0.6474	0.6296
4.9201	9.0	1530	6.4687	0.0027	4.1312	0.9693	0.8814	0.6371	0.6226
4.9201	10.0	1700	6.4563	0.0027	4.0639	0.9701	0.8756	0.6479	0.6197
4.9201	11.0	1870	6.6413	0.0027	4.1177	0.9726	0.8706	0.6255	0.6114
3.7246	12.0	2040	6.7000	0.0027	4.0861	0.9722	0.8851	0.6404	0.6134
3.7246	13.0	2210	7.0551	0.0027	4.2035	0.9714	0.8781	0.6251	0.6089
3.7246	14.0	2380	6.9348	0.0027	4.0477	0.9706	0.8752	0.6354	0.6118
2.6813	15.0	2550	7.0902	0.0027	4.0125	0.9718	0.8756	0.6454	0.6259
2.6813	16.0	2720	7.2237	0.0027	4.0217	0.9685	0.8789	0.6354	0.6031
2.6813	17.0	2890	7.2564	0.0027	4.0108	0.9697	0.8859	0.6383	0.6139
1.921	18.0	3060	7.3150	0.0027	4.0026	0.9701	0.8797	0.6412	0.6172
1.921	19.0	3230	7.3417	0.0027	3.9617	0.9697	0.8801	0.6379	0.6163
1.921	20.0	3400	7.3621	0.0027	3.9585	0.9706	0.8793	0.6400	0.6197

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cpu
• Datasets 4.4.2
• Tokenizers 0.22.1