StanBienaives's picture
Model save
10b8e12 verified
|
raw
history blame
2.43 kB
metadata
license: mit
base_model: flaubert/flaubert_base_cased
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: question_classification
    results: []

question_classification

This model is a fine-tuned version of flaubert/flaubert_base_cased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7638
  • Accuracy: 0.8883
  • F1: 0.8852

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: 2e-05
  • train_batch_size: 8
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss Accuracy F1
No log 1.0 204 2.5153 0.1032 0.0609
No log 2.0 408 2.4600 0.4097 0.2689
2.5497 3.0 612 0.9799 0.5501 0.5544
2.5497 4.0 816 0.5244 0.7994 0.7992
0.8662 5.0 1020 0.5419 0.8281 0.8290
0.8662 6.0 1224 0.4333 0.8510 0.8520
0.8662 7.0 1428 0.4979 0.8711 0.8722
0.2222 8.0 1632 0.5494 0.8625 0.8635
0.2222 9.0 1836 0.6660 0.8739 0.8731
0.0445 10.0 2040 0.6472 0.8825 0.8810
0.0445 11.0 2244 0.7851 0.8940 0.8910
0.0445 12.0 2448 0.7119 0.8854 0.8851
0.0231 13.0 2652 0.8183 0.8883 0.8856
0.0231 14.0 2856 0.7676 0.8883 0.8852
0.0185 15.0 3060 0.7638 0.8883 0.8852

Framework versions

  • Transformers 4.35.2
  • Pytorch 1.13.1+cu117
  • Datasets 2.14.5
  • Tokenizers 0.15.2