Sebastian Lobentanzer
Sebastian Lobentanzer
Provide a better error message for non authenticated access (OpenAI)
It would be nice to be able to benchmark the addition of a few-show prompt in comparison to the naive question, for instance, adding 1-3 example questions and Cypher queries...
Looking towards work such as https://arxiv.org/abs/2310.20440, we could derive benchmarks from the general procedure and the exact issues encountered. E.g., design some artificial texts that are similar to those found...
Since the concrete subtasks including index in the test dataset are only known to the test function, currently the skip happens after the Xinference model has been started up. Moving...
would be nice to have a history of models' performance on tasks
The prototype for cell type annotation is in https://github.com/biocypher/ChatGSE/pull/23 Task: extract generic functionality to biochatter before merging the UI component into ChatGSE There are several other issues that need to...