vit-base-patch16-224-finetuned-barkley

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

• Loss: 0.0036
• Precision: 1.0
• Recall: 1.0
• F1: 1.0
• Accuracy: 1.0
• Top1 Accuracy: 1.0
• Error Rate: 0.0

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: 0.0005
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy	Top1 Accuracy	Error Rate
1.6093	1.0	38	1.4340	0.4769	0.4342	0.4066	0.4149	0.4342	0.5851
1.2908	2.0	76	1.1747	0.6587	0.6118	0.6160	0.6161	0.6118	0.3839
1.0409	3.0	114	0.9174	0.7382	0.7303	0.7293	0.7425	0.7303	0.2575
0.781	4.0	152	0.6528	0.8632	0.8618	0.8622	0.8650	0.8618	0.1350
0.5429	5.0	190	0.4112	0.9417	0.9408	0.9405	0.9443	0.9408	0.0557
0.328	6.0	228	0.2229	0.9809	0.9803	0.9802	0.9811	0.9803	0.0189
0.1837	7.0	266	0.1181	0.9871	0.9868	0.9868	0.9878	0.9868	0.0122
0.1131	8.0	304	0.0680	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0526	9.0	342	0.0387	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0283	10.0	380	0.0328	0.9873	0.9868	0.9869	0.9878	0.9868	0.0122
0.019	11.0	418	0.0224	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.0148	12.0	456	0.0201	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.0095	13.0	494	0.0396	0.9871	0.9868	0.9868	0.9878	0.9868	0.0122
0.007	14.0	532	0.0048	1.0	1.0	1.0	1.0	1.0	0.0
0.011	15.0	570	0.0036	1.0	1.0	1.0	1.0	1.0	0.0
0.0071	16.0	608	0.0092	0.9936	0.9934	0.9934	0.9941	0.9934	0.0059
0.0103	17.0	646	0.0148	0.9936	0.9934	0.9934	0.9944	0.9934	0.0056

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1