cvmil/deit-base-patch16-224_augmented-v2_fft

deit-base-patch16-224_rice-leaf-disease-augmented-v2_fft

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

• Loss: 0.4335
• Accuracy: 0.8810

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: 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7917	1.0	125	1.1504	0.6577
0.6305	2.0	250	0.4746	0.8363
0.1802	3.0	375	0.3663	0.8631
0.0508	4.0	500	0.3550	0.8690
0.0152	5.0	625	0.3373	0.8839
0.0092	6.0	750	0.3433	0.8839
0.0067	7.0	875	0.3768	0.8839
0.002	8.0	1000	0.3861	0.875
0.001	9.0	1125	0.3976	0.8810
0.0009	10.0	1250	0.3989	0.8839
0.0008	11.0	1375	0.4085	0.8839
0.0006	12.0	1500	0.4185	0.8810
0.0004	13.0	1625	0.4294	0.8780
0.0004	14.0	1750	0.4326	0.8780
0.0004	15.0	1875	0.4335	0.8810

Framework versions

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