siglipbase-wikiart-5ep

This model is a fine-tuned version of google/siglip-base-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0301

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-06
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.035	0.1913	500	0.0324
0.029	0.3826	1000	0.0290
0.0291	0.5738	1500	0.0283
0.0239	0.7651	2000	0.0251
0.0232	0.9564	2500	0.0243
0.0175	1.1477	3000	0.0234
0.0163	1.3389	3500	0.0222
0.0164	1.5302	4000	0.0225
0.0168	1.7215	4500	0.0215
0.0155	1.9128	5000	0.0198
0.0094	2.1041	5500	0.0233
0.0087	2.2953	6000	0.0233
0.0093	2.4866	6500	0.0246
0.0093	2.6779	7000	0.0234
0.0087	2.8692	7500	0.0233
0.0058	3.0604	8000	0.0262
0.0063	3.2517	8500	0.0255
0.0059	3.4430	9000	0.0249
0.006	3.6343	9500	0.0254
0.0059	3.8256	10000	0.0262
0.0053	4.0168	10500	0.0283
0.0047	4.2081	11000	0.0290
0.0035	4.3994	11500	0.0286
0.0041	4.5907	12000	0.0305
0.0041	4.7819	12500	0.0302
0.0034	4.9732	13000	0.0301

Framework versions

• Transformers 4.57.1
• Pytorch 2.6.0+cu124
• Datasets 4.4.1
• Tokenizers 0.22.1