metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- imagefolder
metrics:
- accuracy
model-index:
- name: croupier-creature-classifier
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: imagefolder
type: imagefolder
config: alkzar90--croupier-mtg-dataset
split: train
args: alkzar90--croupier-mtg-dataset
metrics:
- name: Accuracy
type: accuracy
value: 0.7058823529411765
croupier-creature-classifier
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.7184
- Accuracy: 0.7059
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.8932 | 1.1 | 100 | 0.9914 | 0.6059 |
0.6608 | 2.2 | 200 | 0.8645 | 0.6588 |
0.6084 | 3.3 | 300 | 0.7326 | 0.7294 |
0.5261 | 4.4 | 400 | 0.7684 | 0.6941 |
0.2511 | 5.49 | 500 | 0.7184 | 0.7059 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.0+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1