Edit model card

finetuned-cards-blackjack

This model is a fine-tuned version of google/vit-base-patch16-224-in21K on the card_images dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5081
  • Accuracy: 0.8696

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: 7
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.3563 0.24 100 1.1495 0.6750
1.3393 0.48 200 1.0388 0.7204
1.2033 0.73 300 0.9324 0.7547
0.9672 0.97 400 0.8558 0.7659
0.8674 1.21 500 0.8456 0.7616
0.8277 1.45 600 0.7563 0.7959
0.8703 1.69 700 0.8465 0.7539
0.893 1.94 800 0.6881 0.8002
0.9454 2.18 900 0.7211 0.8027
0.8109 2.42 1000 0.6369 0.8285
0.8762 2.66 1100 0.6336 0.8396
0.8034 2.91 1200 0.6580 0.8165
0.5833 3.15 1300 0.5828 0.8439
0.8811 3.39 1400 0.6564 0.8259
0.5639 3.63 1500 0.5737 0.8439
0.639 3.87 1600 0.5609 0.8379
0.6455 4.12 1700 0.5820 0.8370
0.5402 4.36 1800 0.5797 0.8345
0.5311 4.6 1900 0.5511 0.8456
0.5734 4.84 2000 0.5444 0.8508
0.5206 5.08 2100 0.5326 0.8636
0.6272 5.33 2200 0.5478 0.8525
0.5124 5.57 2300 0.5296 0.8688
0.5659 5.81 2400 0.5181 0.8705
0.4212 6.05 2500 0.5200 0.8611
0.4338 6.3 2600 0.5135 0.8731
0.3407 6.54 2700 0.5147 0.8722
0.4043 6.78 2800 0.5081 0.8696

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.2.1+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2
Downloads last month
3
Safetensors
Model size
85.8M params
Tensor type
F32
·
Invalid base_model specified in model card metadata. Needs to be a model id from hf.co/models.

Space using Mariofm02/finetuned-cards-blackjack 1