Edit model card

legal-bert-base-uncased-filtered-cuad

This model is a fine-tuned version of nlpaueb/legal-bert-base-uncased on the cuad dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0259

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: 32
  • eval_batch_size: 32
  • 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
0.0394 1.0 31626 0.0265
0.0272 2.0 63252 0.0237
0.021 3.0 94878 0.0259

Framework versions

  • Transformers 4.21.1
  • Pytorch 1.12.0+cu113
  • Datasets 2.4.0
  • Tokenizers 0.12.1
Downloads last month
245
Hosted inference API
Question Answering
Examples
Examples
This model can be loaded on the Inference API on-demand.

Dataset used to train alex-apostolo/legal-bert-base-cuad