legal-luke-base-ner / README.md
lambdavi's picture
Model save
d0df600 verified
metadata
license: apache-2.0
base_model: studio-ousia/luke-base
tags:
  - generated_from_trainer
model-index:
  - name: legal-luke-base-ner
    results: []

legal-luke-base-ner

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

  • Loss: 0.0153
  • F1-type-match: 0.9297
  • F1-partial: 0.9197
  • F1-strict: 0.8794
  • F1-exact: 0.8891

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.06
  • num_epochs: 5

Training results

Training Loss Epoch Step Validation Loss F1-type-match F1-partial F1-strict F1-exact
0.021 1.0 1375 0.0219 0.8297 0.8176 0.7238 0.7525
0.0132 2.0 2750 0.0156 0.8841 0.8722 0.7943 0.8166
0.0087 3.0 4125 0.0155 0.8901 0.8796 0.8271 0.8374
0.0052 4.0 5500 0.0153 0.9190 0.9100 0.8633 0.8750
0.0035 5.0 6875 0.0153 0.9297 0.9197 0.8794 0.8891

Framework versions

  • Transformers 4.36.0
  • Pytorch 2.0.0
  • Datasets 2.17.1
  • Tokenizers 0.15.0