Edit model card

convnext-tiny-224-finetuned-eurosat-albumentations

This model is a fine-tuned version of facebook/convnext-tiny-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3086
  • Accuracy: 0.5

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: 5e-05
  • 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: 50

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 5 2.0663 0.1562
2.0637 2.0 10 2.0571 0.1812
2.0637 3.0 15 2.0389 0.1938
2.0295 4.0 20 2.0109 0.2437
2.0295 5.0 25 1.9754 0.2875
1.9452 6.0 30 1.9335 0.2938
1.9452 7.0 35 1.8869 0.275
1.8289 8.0 40 1.8378 0.2812
1.8289 9.0 45 1.7800 0.3563
1.7075 10.0 50 1.7231 0.3563
1.7075 11.0 55 1.6730 0.3625
1.5909 12.0 60 1.6253 0.3688
1.5909 13.0 65 1.5897 0.3875
1.4997 14.0 70 1.5604 0.4
1.4997 15.0 75 1.5336 0.425
1.4066 16.0 80 1.5147 0.425
1.4066 17.0 85 1.4923 0.425
1.3344 18.0 90 1.4744 0.4375
1.3344 19.0 95 1.4615 0.4437
1.2545 20.0 100 1.4479 0.4437
1.2545 21.0 105 1.4311 0.45
1.1789 22.0 110 1.4222 0.475
1.1789 23.0 115 1.4099 0.4813
1.1186 24.0 120 1.3926 0.4688
1.1186 25.0 125 1.3835 0.4625
1.0685 26.0 130 1.3747 0.4625
1.0685 27.0 135 1.3622 0.4625
0.9935 28.0 140 1.3523 0.4688
0.9935 29.0 145 1.3514 0.45
0.9453 30.0 150 1.3413 0.4688
0.9453 31.0 155 1.3334 0.45
0.9162 32.0 160 1.3239 0.45
0.9162 33.0 165 1.3177 0.475
0.8637 34.0 170 1.3090 0.475
0.8637 35.0 175 1.3078 0.4938
0.8298 36.0 180 1.3086 0.5
0.8298 37.0 185 1.2990 0.5
0.7801 38.0 190 1.2975 0.4938
0.7801 39.0 195 1.2946 0.4938
0.7691 40.0 200 1.2921 0.4875
0.7691 41.0 205 1.2913 0.4938
0.7409 42.0 210 1.2902 0.4875
0.7409 43.0 215 1.2886 0.4875
0.7223 44.0 220 1.2860 0.4938
0.7223 45.0 225 1.2849 0.4875
0.7091 46.0 230 1.2849 0.4875
0.7091 47.0 235 1.2854 0.4875
0.6915 48.0 240 1.2845 0.4875
0.6915 49.0 245 1.2842 0.4875
0.6917 50.0 250 1.2840 0.4875

Framework versions

  • Transformers 4.41.1
  • Pytorch 2.3.0+cu121
  • Datasets 2.19.2
  • Tokenizers 0.19.1
Downloads last month
2
Safetensors
Model size
27.8M params
Tensor type
F32
·
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Finetuned from

Evaluation results