Autocomplete: Add a new 16b exclusive experiment and document existing flags #2632
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
For https://docs.google.com/document/d/1P-Hb8BTbz-vMNFXw3gbAvphArsNvS1_8u4ID-phn8gI/edit, we are considereing moving all traffic to the 16b model. When we first run an experiment for this, the single-line CAR was better with the faster model. Part of this might be due to latency (which will improve with the ST deployment) but parts of this might also just no longer be relevant since we have made so many changes.
This PR adds a new client side feature flag that we can run in the short term to gather data on 16b exclusive setups. It will only have an affect if the user is already inside the
CodyAutocompleteStarCoderHybrid
experiment, which is currently our main StarCoder enablement flag.Test plan
CodyAutocompleteStarCoderHybrid
CodyAutocompleteStarCoder16B