swinv2-tiny-patch4-window8-256-DMAE-15e-6

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 7.9433
  • Accuracy: 0.1087

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: 1.5e-06
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 40

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 0.86 3 7.9433 0.1087
No log 2.0 7 7.9425 0.1087
7.9661 2.86 10 7.9414 0.1087
7.9661 4.0 14 7.9392 0.1087
7.9661 4.86 17 7.9372 0.1087
7.9068 6.0 21 7.9343 0.1087
7.9068 6.86 24 7.9320 0.1087
7.9068 8.0 28 7.9288 0.1087
7.7849 8.86 31 7.9262 0.1087
7.7849 10.0 35 7.9226 0.1087
7.7849 10.86 38 7.9198 0.1087
8.0613 12.0 42 7.9159 0.1087
8.0613 12.86 45 7.9129 0.1087
8.0613 14.0 49 7.9088 0.1087
7.827 14.86 52 7.9057 0.1087
7.827 16.0 56 7.9015 0.1087
7.827 16.86 59 7.8984 0.1087
7.8359 18.0 63 7.8942 0.1087
7.8359 18.86 66 7.8910 0.1087
7.9032 20.0 70 7.8869 0.1087
7.9032 20.86 73 7.8839 0.1087
7.9032 22.0 77 7.8799 0.1087
7.9294 22.86 80 7.8771 0.1087
7.9294 24.0 84 7.8735 0.1087
7.9294 24.86 87 7.8710 0.1087
7.9185 26.0 91 7.8679 0.1087
7.9185 26.86 94 7.8659 0.1087
7.9185 28.0 98 7.8634 0.1087
7.7948 28.86 101 7.8618 0.1087
7.7948 30.0 105 7.8600 0.1087
7.7948 30.86 108 7.8589 0.1087
7.906 32.0 112 7.8578 0.1087
7.906 32.86 115 7.8573 0.1087
7.906 34.0 119 7.8569 0.1087
7.9031 34.29 120 7.8569 0.1087

Framework versions

  • Transformers 4.36.2
  • Pytorch 2.1.2+cu118
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
3
Safetensors
Model size
27.6M params
Tensor type
F32
·
Inference API
Unable to determine this model's library. Check the docs .

Model tree for Augusto777/swinv2-tiny-patch4-window8-256-DMAE-15e-6

Finetuned
(94)
this model

Evaluation results