logo-matching-base

This model is a fine-tuned version of openai/clip-vit-base-patch32 on an unknown dataset. It achieves the following results on the evaluation set:

• Adjusted Rand Score: 0.0388
• Adjusted Mutual Info Score: 0.1580
• Homogeneity Score: 0.6909
• Completeness Score: 0.4506
• Fowlkes Mallows Score: 0.1579
• Pair Confusion Matrix: [[45126, 2524], [12496, 1110]]
• Loss: 0.1711

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: 1e-05
• 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
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Adjusted Rand Score	Adjusted Mutual Info Score	Homogeneity Score	Completeness Score	Fowlkes Mallows Score	Pair Confusion Matrix	Validation Loss
0.0639	1.0	34	0.0132	0.1073	0.6338	0.4284	0.1427	[[44174, 3476], [12476, 1130]]	0.1788
0.0651	2.0	68	0.0190	0.1262	0.6016	0.4393	0.1756	[[42430, 5220], [11902, 1704]]	0.1512
0.0636	3.0	102	-0.0095	0.0882	0.6409	0.4209	0.1097	[[44490, 3160], [12800, 806]]	0.2056
0.0636	4.0	136	0.0474	0.1547	0.7151	0.4503	0.1606	[[45706, 1944], [12586, 1020]]	0.1903
0.0636	5.0	170	0.0469	0.1425	0.6837	0.4430	0.1646	[[45356, 2294], [12484, 1122]]	0.1960
0.0639	6.0	204	0.0251	0.1475	0.7021	0.4457	0.1340	[[45520, 2130], [12752, 854]]	0.1873
0.0643	7.0	238	0.0201	0.1292	0.6920	0.4394	0.1307	[[45340, 2310], [12748, 858]]	0.1720
0.0637	8.0	272	0.0059	0.1170	0.6723	0.4326	0.1163	[[45124, 2526], [12826, 780]]	0.1683
0.0638	9.0	306	0.0230	0.1567	0.6979	0.4515	0.1400	[[45018, 2632], [12624, 982]]	0.1715
0.0637	10.0	340	0.0388	0.1580	0.6909	0.4506	0.1579	[[45126, 2524], [12496, 1110]]	0.1711

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2