metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - alex-apostolo/filtered-cuad
model-index:
  - name: legal-roberta-base-filtered-cuad
    results:
      - task:
          type: question-answering
          name: Question Answering
        dataset:
          name: alex-apostolo/filtered-cuad
          type: alex-apostolo/filtered-cuad
          config: alex-apostolo--filtered-cuad
          split: test
        metrics:
          - type: f1
            value: 71.2057
            name: F1
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNjM4YjdiM2EzZTlhMmExZWY1OTRhYmIzZDkzYjVmNzcyMWY0NmEzZWM1N2JiMTMyMjBlN2JmODc1MmIzMzUyZCIsInZlcnNpb24iOjF9.ytshvfPsqU9mZ4S4h6PpHp8Y72gOg_kuWm4GhlfVTbICzQOZ_dzzgEnytDkmA8hrHGQb7o2xZWR5At0b1iv8Aw
          - type: exact
            value: 69.0171
            name: Exact Match
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMWI4M2Y0Y2FmNjA4Mjg5ZDk3M2ZkMmM1YzM3Y2MyYWJkMTgzZGYwZWM1ZTNiYWU2NTg2NDU4OGNkYjEzYTY4NSIsInZlcnNpb24iOjF9.RKnHBA-PIUGveGyhZyQBNUU2D8nbNCR11pCtwu57Xua9BlpZAqFVuvXGrj26okArrvRtTaAfyRg2BKNn15KlBg
          - type: loss
            value: 0.06303470581769943
            name: loss
            verified: true
            verifyToken: >-
              eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNWI1NTEyZDBhMjVjMjNjM2E1NTUzNjU0MjA3ZDg2ZDc4YTkxZjQ1OTE5NDYxYWRkODZkZDczNmMyYjg1OWQ2NiIsInZlcnNpb24iOjF9.hbDPd4aefzbVUu1LWAUAaIDJTCzjrpBvmPqyHlzz0bwqAxeV6t6poAsb0Pt6jRkNkRhFR2hM_vlUGfzQbH-RDQ

legal-roberta-base-filtered-cuad

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

Loss: 0.0428

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: 22
eval_batch_size: 22
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
0.0556	1.0	12279	0.0517
0.0406	2.0	24558	0.0425
0.0332	3.0	36837	0.0428

Framework versions

Transformers 4.21.1
Pytorch 1.12.0+cu113
Datasets 2.4.0
Tokenizers 0.12.1