Edit model card

convnext-tiny-224-finetuned-papsmear

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: 0.4893
  • Accuracy: 0.8088

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.706 0.9935 38 1.6028 0.2794
1.3867 1.9869 76 1.2961 0.4853
1.0784 2.9804 114 1.0588 0.5221
0.9128 4.0 153 0.8886 0.6618
0.7466 4.9935 191 0.8913 0.6029
0.6886 5.9869 229 0.7380 0.7059
0.6198 6.9804 267 0.7622 0.7132
0.6001 8.0 306 0.7083 0.6838
0.5542 8.9935 344 0.5909 0.7721
0.5161 9.9869 382 0.5909 0.7574
0.4631 10.9804 420 0.5677 0.7721
0.4284 12.0 459 0.5229 0.7868
0.4334 12.9935 497 0.5160 0.8015
0.4386 13.9869 535 0.4788 0.8015
0.3728 14.9020 570 0.4893 0.8088

Framework versions

  • Transformers 4.44.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.19.1
Downloads last month
10
Safetensors
Model size
27.8M params
Tensor type
F32
·
Inference Examples
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.

Model tree for rukundob451/convnext-tiny-224-finetuned-papsmear

Finetuned
(28)
this model

Evaluation results