Christoph Sander
Christoph Sander
I have to say though that some tweets happen to appear twice in the all feed. But that's not a problem for me. Thanks for the fix anyway
I guess the issue is sequence labeling vs. text classification...yet, I was wondering if NER training can be done via a CSV file as well instead of BIO.
This sounds awesome! Thanks! It would be promising to also involve https://github.com/CIRCSE and their many efforts related to the [LiLa project](https://lila-erc.eu/) @passarom. If I'm right, they also included more Medieval...
Great! As for the visual part and the UI: I fully trust in your and others' expertise! For the logic: If I am not mistaken, narrowing down the tags after...
a file permission issue. The oxigraph server cannot run in ro-mode
Thanks, that solved the initial problem! I would like to use the bulk load from pyoxigraph (as working in a flask environment) and have the API endpoint as a read-only...
Fantastic: `?default` did the trick -- although for trig upload I get a `No supported content Content-Type given: application/trig`. Not a problem. pyoxigraph's `bulk_load` takes trig but I havent't been...
I confirm, the `serve-secondary` works but only took 1088521 of the total of 1225120 triples before a restart of the service. **Is there any way to force a reload of...
I'll try to reproduce and if persistent I will let you know
would be an easy fix, thanks to `use_default_graph_as_union`: ```python query = """ SELECT ?s ?p ?o ?a WHERE { ?s ?p ?o . } LIMIT 10 """ results = store.query(query=query,use_default_graph_as_union=True)...