Edit model card

mit-b2-finetuned-memes

This model is a fine-tuned version of aaraki/vit-base-patch16-224-in21k-finetuned-cifar10 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4137
  • Accuracy: 0.8524

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: 0.00012
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • 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
0.9727 0.99 40 0.8400 0.7334
0.5305 1.99 80 0.5147 0.8284
0.3124 2.99 120 0.4698 0.8145
0.2263 3.99 160 0.3892 0.8563
0.1453 4.99 200 0.3874 0.8570
0.1255 5.99 240 0.4097 0.8470
0.0989 6.99 280 0.3860 0.8570
0.0755 7.99 320 0.4141 0.8539
0.08 8.99 360 0.4049 0.8594
0.0639 9.99 400 0.4137 0.8524

Framework versions

  • Transformers 4.22.1
  • Pytorch 1.12.1+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
10

Evaluation results