Raphael Sourty

Hi @artyomche9 , Thanks for your feedback, the library does not allow for the moment to easily add new entities / relations to a model already trained. For the moment...

Could you give me a use case where you need to remove knowledge? I don't have a clear idea about the usefulness of removing knowledge yet

I think updating knowledge is better than removing knowledge. We could ask the model to make Istanbul no longer the capital of Turkey, by reducing the score but I think...

If we are updating both embeddings of Ankara and Turquie, Ankara would be more the capital of Turkey than Istanbul. At least I hope so.

Hi @raul-parada, The perceptron has a ´predict_proba_many´ method in order to return probabilities over a batch of input samples. Is it what you are looking for? As part of River,...

Seems like it's a problem that's come from your input that does not fit River api. When using learn_one and predict_one, your inpute features (x) should be formated as {feature_0:...

Hi @Asynchro-Epool, did not know about [notion-py](https://github.com/jamalex/notion-py) which would perfectly fit for this tool, I'll add this to my to-do list (of course contributions are welcome). I'll try to think...

Hi @richardklafter, I used the msmarco dataset, trained using neural-search without negative samples. The checkpoint is either a distilbert or a co-condenser, don't remember. I trained the model with a...

Thank you @arthur-75 for this MR, the best I think would be to add an `index` directory with a file `annoy.py` in this directory. The class would be `Annoy()` with...

Hi @amallia, that would be very relevant. If you're interested in developping it, any contribution is welcome. Once I have more time, I'll study this :) The code would rely...