Edit model card

DBERT_Emotions_tuned

This model is a fine-tuned version of distilbert-base-uncased on the emotion dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1828
  • Accuracy: 0.925

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: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 0.1 100 0.7513 0.7365
No log 0.2 200 0.3693 0.8895
No log 0.3 300 0.3118 0.906
No log 0.4 400 0.3048 0.9055
0.5368 0.5 500 0.2649 0.9225
0.5368 0.6 600 0.2192 0.9235
0.5368 0.7 700 0.2254 0.9245
0.5368 0.8 800 0.2016 0.931
0.5368 0.9 900 0.1685 0.935
0.2254 1.0 1000 0.1926 0.9295
0.2254 1.1 1100 0.2128 0.928
0.2254 1.2 1200 0.2008 0.9325
0.2254 1.3 1300 0.1662 0.9385
0.2254 1.4 1400 0.1945 0.939
0.1315 1.5 1500 0.1652 0.939
0.1315 1.6 1600 0.1820 0.938
0.1315 1.7 1700 0.1660 0.938
0.1315 1.8 1800 0.1590 0.93
0.1315 1.9 1900 0.1601 0.935
0.1295 2.0 2000 0.1645 0.9345
0.1295 2.1 2100 0.1845 0.9305
0.1295 2.2 2200 0.1784 0.9355
0.1295 2.3 2300 0.2042 0.9365
0.1295 2.4 2400 0.1852 0.9365
0.0891 2.5 2500 0.1797 0.94
0.0891 2.6 2600 0.1741 0.9365
0.0891 2.7 2700 0.1758 0.9385
0.0891 2.8 2800 0.1771 0.944
0.0891 2.9 2900 0.1688 0.9385
0.0848 3.0 3000 0.1671 0.94

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2
Downloads last month
13
Safetensors
Model size
67M params
Tensor type
F32
·

Finetuned from

Dataset used to train NPCProgrammer/DBERT_Emotions_tuned

Evaluation results