swinv2-plantclef

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

• Loss: 1.0548
• Accuracy: 0.8199
• F1: 0.8190

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use 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: 16
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.1414	1.0	897	0.9819	0.7171	0.7046
0.654	2.0	1794	0.7608	0.7694	0.7688
0.394	3.0	2691	0.7461	0.7795	0.7767
0.2437	4.0	3588	0.7369	0.7917	0.7908
0.1428	5.0	4485	0.7939	0.7945	0.7929
0.0878	6.0	5382	0.8352	0.7958	0.7950
0.0621	7.0	6279	0.8802	0.7945	0.7928
0.0353	8.0	7176	0.9028	0.8011	0.8005
0.0241	9.0	8073	0.9592	0.8043	0.8045
0.0241	10.0	8970	1.0075	0.8068	0.8047
0.0129	11.0	9867	1.0254	0.8127	0.8120
0.0058	12.0	10764	1.0340	0.8162	0.8151
0.007	13.0	11661	1.0661	0.8165	0.8159
0.0052	14.0	12558	1.0533	0.8168	0.8166
0.0049	15.0	13455	1.0660	0.8174	0.8164
0.015	16.0	14352	1.0548	0.8199	0.8190

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.0
• Datasets 3.1.0
• Tokenizers 0.20.1