Edit model card

distilbert-base-uncased-finetuned-emotion

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: 0.1538
  • Accuracy: 0.936
  • F1: 0.9361

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: 256
  • eval_batch_size: 256
  • 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 63 0.8113 0.745 0.6916
No log 2.0 126 0.2909 0.911 0.9107
No log 3.0 189 0.2003 0.928 0.9277
0.5629 4.0 252 0.1698 0.938 0.9376
0.5629 5.0 315 0.1561 0.9365 0.9364
0.5629 6.0 378 0.1531 0.933 0.9334
0.5629 7.0 441 0.1584 0.9355 0.9345
0.1065 8.0 504 0.1493 0.9325 0.9321
0.1065 9.0 567 0.1504 0.936 0.9364
0.1065 10.0 630 0.1481 0.9395 0.9395
0.1065 11.0 693 0.1501 0.935 0.9353
0.0684 12.0 756 0.1504 0.936 0.9360
0.0684 13.0 819 0.1526 0.935 0.9352
0.0684 14.0 882 0.1526 0.9355 0.9357
0.0684 15.0 945 0.1538 0.936 0.9361

Framework versions

  • Transformers 4.37.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.14.6
  • Tokenizers 0.15.2
Downloads last month
3
Safetensors
Model size
67M params
Tensor type
F32
·

Finetuned from

Dataset used to train LeoTungAnh/distilbert-base-uncased-finetuned-emotion

Evaluation results