djbp's picture
End of training
847b644 verified
|
raw
history blame
2.46 kB
metadata
license: apache-2.0
base_model: microsoft/swin-tiny-patch4-window7-224
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: swin-tiny-patch4-window7-224-MM_Classification
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: train
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8915401301518439

swin-tiny-patch4-window7-224-MM_Classification

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2895
  • Accuracy: 0.8915

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: 128
  • eval_batch_size: 128
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 512
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.0635 0.9846 16 0.7524 0.6725
0.4571 1.9692 32 0.3692 0.8742
0.3819 2.9538 48 0.3500 0.8688
0.3278 4.0 65 0.3158 0.8796
0.2941 4.9846 81 0.2886 0.8883
0.2912 5.9692 97 0.2895 0.8915
0.2575 6.9538 113 0.2801 0.8839
0.2604 8.0 130 0.2847 0.8861
0.2519 8.9846 146 0.2804 0.8872
0.2592 9.8462 160 0.2795 0.8872

Framework versions

  • Transformers 4.43.2
  • Pytorch 1.13.1+cu117
  • Datasets 2.20.0
  • Tokenizers 0.19.1