test_swin-base-patch4-window7-224_rice-leaf-disease-augmented-v3_tl

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5013
• Accuracy: 0.8143

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: 0.0003
• train_batch_size: 64
• eval_batch_size: 64
• 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: cosine_with_restarts
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8829	1.0	128	1.4836	0.5765
1.2287	2.0	256	0.9555	0.7068
0.8672	3.0	384	0.7629	0.7492
0.7193	4.0	512	0.6680	0.7883
0.6461	5.0	640	0.6311	0.7850
0.6029	6.0	768	0.6150	0.7883
0.5843	7.0	896	0.6107	0.7883
0.5781	8.0	1024	0.6103	0.7883
0.5642	9.0	1152	0.5867	0.7948
0.5154	10.0	1280	0.5532	0.7915
0.4882	11.0	1408	0.5411	0.8111
0.4693	12.0	1536	0.5354	0.8111
0.4604	13.0	1664	0.5351	0.8078
0.4557	14.0	1792	0.5344	0.8078
0.4502	15.0	1920	0.5257	0.8111
0.4257	16.0	2048	0.5174	0.8111
0.4077	17.0	2176	0.5151	0.8013
0.3969	18.0	2304	0.4979	0.8176
0.3904	19.0	2432	0.5003	0.8143
0.3861	20.0	2560	0.5013	0.8143

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0