As a model admin, I want to be able to collect the feedback on model responses and use it to direct model finetuning.
Consider and review suitable frameworks such as https://langfuse.com/ which provide standardized scoring and evalution of model responses. The content should be usable in a review process that would rely on instruction tuning (possibly using a low code approach such as InstructLab integrated in a User Interface)
As a model admin, I want to be able to collect the feedback on model responses and use it to direct model finetuning.
Consider and review suitable frameworks such as https://langfuse.com/ which provide standardized scoring and evalution of model responses. The content should be usable in a review process that would rely on instruction tuning (possibly using a low code approach such as InstructLab integrated in a User Interface)