Edit model card

convnextv2-tiny-1k-224-finetuned-cassava-leaf-disease

This model is a fine-tuned version of facebook/convnextv2-tiny-1k-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4109
  • Accuracy: 0.8650

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: 480
  • eval_batch_size: 480
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 1920
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Accuracy
7.8796 0.98 10 3.9572 0.1706
2.3762 1.95 20 1.4334 0.6178
1.1413 2.93 30 0.8877 0.6841
0.7549 4.0 41 0.6403 0.7724
0.5904 4.98 51 0.5366 0.8098
0.5152 5.95 61 0.4799 0.8369
0.4764 6.93 71 0.4567 0.8486
0.4386 8.0 82 0.4421 0.8509
0.4306 8.98 92 0.4381 0.8519
0.4266 9.95 102 0.4296 0.8603
0.4072 10.93 112 0.4196 0.8593
0.4033 12.0 123 0.4127 0.8621
0.3982 12.98 133 0.4125 0.8640
0.3993 13.95 143 0.4097 0.8631
0.3812 14.63 150 0.4109 0.8650

Framework versions

  • Transformers 4.37.2
  • Pytorch 2.2.1
  • Datasets 2.18.0
  • Tokenizers 0.15.1
Downloads last month
12
Safetensors
Model size
28.6M params
Tensor type
F32
·

Finetuned from

Evaluation results