swinv2-tiny-patch4-window8-256-DMAE-U3

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: 1.7136
  • Accuracy: 0.4565

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: 5.5e-05
  • 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: 0.05
  • num_epochs: 40

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 0.86 3 7.9328 0.1087
No log 2.0 7 7.7596 0.1087
7.8559 2.86 10 7.3241 0.1087
7.8559 4.0 14 6.4450 0.1087
7.8559 4.86 17 5.7723 0.1087
6.3363 6.0 21 4.8772 0.1087
6.3363 6.86 24 4.2678 0.1087
6.3363 8.0 28 3.5000 0.1087
4.1887 8.86 31 2.9766 0.1087
4.1887 10.0 35 2.3876 0.1087
4.1887 10.86 38 2.0429 0.1087
2.602 12.0 42 1.7136 0.4565
2.602 12.86 45 1.5478 0.4565
2.602 14.0 49 1.3874 0.4565
1.6353 14.86 52 1.2968 0.4565
1.6353 16.0 56 1.2225 0.4565
1.6353 16.86 59 1.2071 0.4565
1.2533 18.0 63 1.2177 0.3261
1.2533 18.86 66 1.2197 0.3261
1.2088 20.0 70 1.2112 0.4565
1.2088 20.86 73 1.2101 0.4565
1.2088 22.0 77 1.2092 0.4565
1.1798 22.86 80 1.2081 0.4565
1.1798 24.0 84 1.2076 0.4565
1.1798 24.86 87 1.2049 0.4565
1.1825 26.0 91 1.2045 0.4565
1.1825 26.86 94 1.2029 0.4565
1.1825 28.0 98 1.2022 0.4565
1.1943 28.86 101 1.2014 0.4565
1.1943 30.0 105 1.2040 0.4565
1.1943 30.86 108 1.2050 0.4565
1.1772 32.0 112 1.2031 0.4565
1.1772 32.86 115 1.2019 0.4565
1.1772 34.0 119 1.2013 0.4565
1.1945 34.29 120 1.2012 0.4565

Framework versions

  • Transformers 4.36.2
  • Pytorch 2.1.2+cu118
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
5
Safetensors
Model size
27.6M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no library tag.

Model tree for Augusto777/swinv2-tiny-patch4-window8-256-DMAE-U3

Finetuned
(121)
this model

Evaluation results