predict-plates / README.md
data-silence's picture
Update README.md
e9ca70f verified
metadata
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: results
    results: []
library_name: transformers
pipeline_tag: image-classification

results

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0398
  • Accuracy: 1.0

Model description

This model was trained for the Kaggle competition Cleaned vs Dirty V2. Despite good results in training, the model shows poor results on test data, and should not be used in this competition.

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 20 0.0907 1.0
No log 2.0 40 0.0468 1.0
No log 3.0 60 0.0398 1.0

Framework versions

  • Transformers 4.42.4
  • Pytorch 2.3.1+cu121
  • Datasets 2.21.0
  • Tokenizers 0.19.1