job G B E R T

Introduction

JOBGBERT is a domain-adapted, transformer-based language model designed to process German-speaking job advertisements. It is built upon the deepset/gbert-base model and further trained with 4 million job ads from Switzerland, spanning the years 1990 to 2020.

Architecture

• Model Type: BERT base
• Language: German
• Domain: Job advertisements

Training

JOBGBERT was adapted to its specific domain through continued in-domain pretraining using a large dataset of job advertisements. This dataset consists of 5.9 GB of text extracted from Swiss job advertisements over three decades.

Guide: Running Locally

1. Installation: Ensure you have Python and PyTorch installed. Use the Hugging Face Transformers library to load the model.
2. Loading the Model: Use the from_pretrained method to load agne/jobGBERT.
3. Inference: Implement the fill-mask pipeline for masked language modeling tasks.

For optimal performance, consider using cloud GPU services like AWS, Google Cloud, or Azure.

License

JOBGBERT is distributed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (cc-by-nc-sa-4.0). Proper citation is required when using this model in academic or research contexts.