-
Is their a possibility to get suggestions for terms that are important in a document or a set of documents, but that are not present in the supported vocabulary?
-
I'm trying to adapt Omikuji (via the integration with [Annif](https://github.com/NatLibFi/Annif)) into a [Data Version Control](https://dvc.org/) workflow. In DVC, large model files are typically stor…
-
I observed a dependency mismatch in `annif==0.54.*` In fact I got this also for older versions.
This leads to an error exit code when using `pipenv`. I think you don't support `pipenv` at the moment.…
mo-fu updated
3 years ago
-
**Describe the bug**
A clear and concise description of the bug.
We've been having issues with our CI in GitHub Actions for the last few hours, and think it might be because of Coverage 6.3 - it's…
-
Hi,
to update an existing vocabulary with more than 1.3 million of concepts takes hours, because the existings graph is loaded and each triple is checked. I think it would be an good improvement to…
-
I tried to use the learn-method with nn backend. Backend consists ready-made Epikurous backends (mllm, fasttext, bonsai).
I use the docker version.
nn was trained with tsv-file with 4384 short …
-
When running evaluations for new models with four parallel jobs I noticed that evaluation of nn_ensemble projects never finished. Without multiprocessing there was no problem.
Minimal example:
``…
-
Training the MLLM backend can be a bit slow. Most of the time is spent generating candidates from the training documents. This could probably be done faster by using parallel processing. It's noted as…
-
I have an nn-ensemble model using Fasttext, Omikuji, and MLLM as backends. The model works as expected for suggestions, including with the rest API. I was testing the online learning API, and the back…
-
Looking at the documentation of the access settings, it seems that the purpose of hidden projects is to hide them from the UI, but still expose them to the API, and the purpose of private projects is …