-
Couldn't load subscription status.
- Fork 25.6k
Open
Labels
:mlMachine learningMachine learning:ml/ChunkingFeature:GenAIFeatures around GenAIFeatures around GenAIFeature:NLPFeatures and issues around NLPFeatures and issues around NLPTeam:MLMeta label for the ML teamMeta label for the ML team
Description
The purpose of this issue is to add configuration parameters to this PR to only send the first N chunks generated to the rerank model. After adding this functionality we will need to develop new evaluation scripts to test the impact of various values of N against cost/latency/rerank accuracy.
Metadata
Metadata
Assignees
Labels
:mlMachine learningMachine learning:ml/ChunkingFeature:GenAIFeatures around GenAIFeatures around GenAIFeature:NLPFeatures and issues around NLPFeatures and issues around NLPTeam:MLMeta label for the ML teamMeta label for the ML team