cvmil/vit-base-patch16-224_augmented-v2_tl

vit-base-patch16-224_rice-leaf-disease-augmented-v2_tl

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

• Loss: 0.6919
• Accuracy: 0.7679

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
2.1171	1.0	63	1.8775	0.2946
1.6139	2.0	126	1.3619	0.5476
1.1727	3.0	189	1.1003	0.6577
0.9586	4.0	252	0.9665	0.7232
0.8409	5.0	315	0.8663	0.7440
0.7632	6.0	378	0.8322	0.7381
0.7093	7.0	441	0.8039	0.7470
0.6667	8.0	504	0.7722	0.75
0.6353	9.0	567	0.7477	0.7560
0.6101	10.0	630	0.7304	0.7589
0.5894	11.0	693	0.7229	0.7649
0.5737	12.0	756	0.7130	0.7619
0.5627	13.0	819	0.7033	0.7649
0.5524	14.0	882	0.7009	0.7649
0.5439	15.0	945	0.6945	0.7679
0.5397	16.0	1008	0.6937	0.7649
0.5357	17.0	1071	0.6933	0.7679
0.5337	18.0	1134	0.6919	0.7679
0.5322	19.0	1197	0.6921	0.7679
0.5325	20.0	1260	0.6919	0.7679

Framework versions

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