-
Some ideas as to how the general public can contribute to the scoring whilst preventing the ability of bad faith actors to skew results in favour of non-reputable sources. n.b These are not all my ide…
-
I feel like the current one works out of sheer luck, I would love to have some assistance from someone with a very strong mathematical background to help me figure out the best static evaluator.
-
-
According to https://scikit-learn.org/stable/modules/grid_search.html,
_HalvingRandomSearchCV and HalvingGridSearchCV do not support multimetric scoring._
When will this be implemented?
When…
-
I would be very nice if you could push the most important score metrics, the current score and performance to prometheus.
We are then able to analyze the data in Grafana.
-
Even though this seems very unlikely, for the multiple choice task the models might return the same scores for some of the options. If these options have the highest score, and the reference label is …
-
It seems as though having duplicate tokens in a name is causing elasticsearch to score the result higher (this is due to how the TF/IDF scoring works)
While it's impossible to have exact duplicate …
-
It might be useful to take the creation dates of documents into account in our scoring scheme. The information can be taken either from the mtime of file systems or the `CreationDate` field of PDF met…
-
I get to know that delete is not happening on hybrid index. Deleting the embeddings using its "id" is only deleting the embedding list but the parallely created "scoring.terms" and "scoring" remains u…
-
The current stringify-then-regex-extract approach is kind of horrifying. ~I must have been on a regex kick when I wrote it.~ But, it was the best way I could think of at the time to retain each object…