swinv2-tiny-patch4-window16-256-finetuned-tekno24

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window16-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2871
• Accuracy: 0.4224
• F1: 0.3135
• Precision: 0.4313
• Recall: 0.4224

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-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.3705	0.9963	68	1.3584	0.3563	0.2590	0.2886	0.3563
1.3515	1.9927	136	1.3392	0.3921	0.2606	0.3925	0.3921
1.3498	2.9890	204	1.3301	0.3912	0.2501	0.4247	0.3912
1.3351	4.0	273	1.3225	0.3930	0.2452	0.5371	0.3930
1.3212	4.9963	341	1.3128	0.3949	0.2556	0.4641	0.3949
1.3316	5.9927	409	1.3052	0.4004	0.2723	0.4129	0.4004
1.3269	6.9890	477	1.2980	0.4068	0.2850	0.4305	0.4068
1.3034	8.0	546	1.2927	0.4123	0.2924	0.4448	0.4123
1.3165	8.9963	614	1.2884	0.4215	0.3096	0.4453	0.4215
1.3306	9.9634	680	1.2871	0.4224	0.3135	0.4313	0.4224

Framework versions

• Transformers 4.42.4
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1