cvmil/dinov2-base_augmented-v2_tl

dinov2-base_rice-leaf-disease-augmented-v2_tl

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: 0.5013
• Accuracy: 0.8512

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: 128
• eval_batch_size: 128
• 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
2.0771	1.0	63	1.5173	0.4881
1.0706	2.0	126	0.9022	0.7173
0.6568	3.0	189	0.7203	0.7619
0.5123	4.0	252	0.6513	0.7827
0.4311	5.0	315	0.5929	0.8125
0.3812	6.0	378	0.5802	0.8274
0.3426	7.0	441	0.5669	0.8274
0.3145	8.0	504	0.5452	0.8333
0.2919	9.0	567	0.5532	0.8155
0.276	10.0	630	0.5275	0.8423
0.263	11.0	693	0.5189	0.8512
0.2502	12.0	756	0.5181	0.8512
0.2416	13.0	819	0.5058	0.8482
0.2343	14.0	882	0.5050	0.8542
0.2292	15.0	945	0.5009	0.8482
0.2245	16.0	1008	0.5057	0.8482
0.2224	17.0	1071	0.5040	0.8512
0.2205	18.0	1134	0.5021	0.8482
0.2195	19.0	1197	0.5017	0.8512
0.2188	20.0	1260	0.5013	0.8512

Framework versions

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