Prot10's picture
End of training
285604a
metadata
license: apache-2.0
base_model: facebook/convnextv2-base-1k-224
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: convnextv2-base-1k-224-for-pre_evaluation
    results: []

convnextv2-base-1k-224-for-pre_evaluation

This model is a fine-tuned version of facebook/convnextv2-base-1k-224 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.3599
  • Accuracy: 0.4190

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: 30

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.6 1.0 16 1.5316 0.2961
1.5084 2.0 32 1.5061 0.2849
1.5134 3.0 48 1.4968 0.3240
1.4663 4.0 64 1.4607 0.3352
1.4046 5.0 80 1.4509 0.3268
1.4085 6.0 96 1.4423 0.3883
1.3443 7.0 112 1.4005 0.4022
1.3025 8.0 128 1.3599 0.4190
1.2627 9.0 144 1.3638 0.3911
1.2099 10.0 160 1.4058 0.3492
1.2086 11.0 176 1.4431 0.3408
1.1393 12.0 192 1.4143 0.3492
1.1039 13.0 208 1.4305 0.3883
1.0551 14.0 224 1.5203 0.3520
1.0368 15.0 240 1.5117 0.3324
0.9753 16.0 256 1.4545 0.3771
0.938 17.0 272 1.5396 0.3352
0.899 18.0 288 1.5770 0.3408
0.8629 19.0 304 1.7106 0.3128
0.8674 20.0 320 1.5864 0.3352
0.7789 21.0 336 1.6129 0.3408
0.7426 22.0 352 1.6353 0.3603
0.7677 23.0 368 1.6793 0.3464
0.7172 24.0 384 1.6759 0.3575
0.6809 25.0 400 1.7013 0.3659
0.6619 26.0 416 1.7108 0.3631
0.6656 27.0 432 1.7327 0.3715
0.6258 28.0 448 1.7378 0.3547
0.6173 29.0 464 1.7461 0.3603
0.6214 30.0 480 1.7475 0.3520

Framework versions

  • Transformers 4.33.1
  • Pytorch 2.0.1+cu118
  • Datasets 2.14.5
  • Tokenizers 0.13.3