dinov2-base_rice-leaf-disease-augmented-v2_test_fft

This model is a fine-tuned version of facebook/dinov2-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0568
• Accuracy: 0.8036

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: 0.001
• 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: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
2.973	1.0	125	0.2054	2.1679
1.6589	2.0	250	0.5089	1.3584
1.4112	3.0	375	0.5595	1.2409
1.3047	4.0	500	0.5655	1.2476
1.2215	5.0	625	0.6458	1.0247
1.0558	6.0	750	0.6607	0.9527
0.9794	7.0	875	0.6935	0.8817
0.8155	8.0	1000	0.7262	0.8064
0.6318	9.0	1125	0.7381	0.7826
0.4447	10.0	1250	0.7738	0.7427
0.713	11.0	1375	0.8410	0.7173
0.6409	12.0	1500	0.7657	0.7411
0.5784	13.0	1625	0.8426	0.7262
0.4593	14.0	1750	0.7595	0.7708
0.3596	15.0	1875	0.8236	0.7679
0.2488	16.0	2000	0.8065	0.7798
0.1765	17.0	2125	0.7864	0.8244
0.1058	18.0	2250	0.8167	0.8214
0.0482	19.0	2375	0.9683	0.8125
0.0144	20.0	2500	1.0568	0.8036

Framework versions

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