Instructions to use Christoph911/GELECTRA-distilled-LegalQuAD with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Christoph911/GELECTRA-distilled-LegalQuAD with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("question-answering", model="Christoph911/GELECTRA-distilled-LegalQuAD")# Load model directly from transformers import AutoTokenizer, AutoModelForQuestionAnswering tokenizer = AutoTokenizer.from_pretrained("Christoph911/GELECTRA-distilled-LegalQuAD") model = AutoModelForQuestionAnswering.from_pretrained("Christoph911/GELECTRA-distilled-LegalQuAD") - Notebooks
- Google Colab
- Kaggle
YAML Metadata Error:"widget[0].text" is not allowed to be empty
GELECTRA-distilled-LegalQuAD
Overview
Language model: GELECTRA-distilled
Language: German
Downstream-task: Extractive QA
Training data: German-legal-SQuAD
Eval data: German-legal-SQuAD testset
Hyperparameters
batch_size = 10
n_epochs = 2
max_seq_len=256,
learning_rate=1e-5,
## Eval results
Evaluated on German-legal-SQuAD testset
"exact": 34.294
"f1": 64.991
- Downloads last month
- 6