sentiment-unipelt / README.md
apwic's picture
End of training
7ec98c8 verified
metadata
language:
  - id
license: mit
base_model: indolem/indobert-base-uncased
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: sentiment-unipelt
    results: []

sentiment-unipelt

This model is a fine-tuned version of indolem/indobert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2928
  • Accuracy: 0.9023
  • Precision: 0.8842
  • Recall: 0.8783
  • F1: 0.8812

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: 30
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 20.0

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1
0.5535 1.0 122 0.4992 0.7293 0.6646 0.6285 0.6373
0.444 2.0 244 0.4053 0.8170 0.7847 0.8256 0.7961
0.3464 3.0 366 0.3425 0.8421 0.8345 0.7683 0.7905
0.2852 4.0 488 0.3136 0.8722 0.8445 0.8496 0.8470
0.2608 5.0 610 0.3060 0.8722 0.8445 0.8496 0.8470
0.2415 6.0 732 0.3100 0.8647 0.8325 0.8642 0.8447
0.2329 7.0 854 0.2860 0.8847 0.8567 0.8734 0.8642
0.199 8.0 976 0.2879 0.8872 0.8672 0.8577 0.8622
0.1939 9.0 1098 0.2826 0.8897 0.8659 0.8695 0.8676
0.1806 10.0 1220 0.2982 0.8797 0.8795 0.8224 0.8439
0.1674 11.0 1342 0.2735 0.8947 0.8730 0.8730 0.8730
0.1553 12.0 1464 0.2753 0.8947 0.8757 0.8680 0.8717
0.1431 13.0 1586 0.2937 0.8922 0.8785 0.8562 0.8662
0.1417 14.0 1708 0.2911 0.9073 0.8823 0.9019 0.8910
0.1236 15.0 1830 0.2956 0.9023 0.8828 0.8808 0.8818
0.1304 16.0 1952 0.3011 0.9023 0.8773 0.8933 0.8846
0.1164 17.0 2074 0.2943 0.8997 0.8778 0.8816 0.8797
0.1144 18.0 2196 0.2937 0.8972 0.8732 0.8823 0.8776
0.1198 19.0 2318 0.2985 0.8972 0.8812 0.8673 0.8738
0.1104 20.0 2440 0.2928 0.9023 0.8842 0.8783 0.8812

Framework versions

  • Transformers 4.39.3
  • Pytorch 2.3.0+cu121
  • Datasets 2.19.1
  • Tokenizers 0.15.2