malteos
malteos
Thanks for the amazing work! Can you provide pretrained dumps as download? That would make it much easier to test everything locally without the training step.
Hi, thanks for your work on this project! I was surprised the see that your perplexity implementation uses the base two exponential. See https://github.com/huggingface/evaluate/blob/main/metrics/perplexity/perplexity.py#L183 Is this indented or a bug?
Do you plan to release the model weights of German version as well?
See - https://github.com/laws-africa/indigo - https://www.mdpi.com/2078-2489/11/1/10
Task: Provide a more verbose explanation of the processing pipeline and how the results are used in the front-end. https://oldp.readthedocs.io/en/latest/processing.html - Define input: HTML / plain text etc. - How...
Use multiple threads for some processing tasks - extract refs etc. See CourtListner - http://docs.celeryproject.org/en/latest/index.html
- Use https://github.com/xcash/bootstrap-autocomplete - Prepare list with most popular keywords, existing tags, ..., and curated list of user-generated search terms.
See - Blogpost https://www.elastic.co/blog/text-similarity-search-with-vectors-in-elasticsearch - Available as Docker image: https://hub.docker.com/_/elasticsearch
Flow: 1. GitHub triggers blog app 2. checkout latest git 3. generate static site (www is hosted by NGINX + blog) 4. include blog posts by language in Django
Extending OLDP to other countries would get much easier if we support themes. - [x] Override homepage + header + footer with template Germany-specific elements should be moved to a...