Proof-of-concept: on-device LLM generated nudges #94
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Collaborator
Author
|
@PSchmiedmayer @lukaskollmer I've gotten a demo/debug view of on-device LLM generation of nudges working! Would greatly appreciate a code review. Thanks for your time! |
Collaborator
Author
|
It seems that the CodeQL tests are failing because Metal Toolchain is missing. Should it be added to the test harness? |
awangwang
force-pushed
the
awangwang/on-device-LLMs-refactored
branch
from
2569ba7 to
967b775
Compare
Collaborator
Author
|
@PSchmiedmayer @lukaskollmer would appreciate your help with a code review this week, especially if you think the PR is not ready to be pushed so I work on those changes! Thanks! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Proof-of-concept: on-device LLM generated nudges
♻️ Current situation & Problem
MHC is in need of an implementation of on-device execution of LLMs in order to generate personalized motivational notifications for users.
⚙️ Release Notes
LLMRunnerandOnDeviceNudgeServicetoMyHeartCountsDelegate.swift.downloadLLMfeature flag inFeatureFlags.swift.OnDeviceNudgeService.swifthandles operations to retrieve user account data and then build up an LLM prompt customized using the user data. The prompt asks the LLM to create 7 motivational messages (nudges) intended to be sent to the user as a daily notification. It also allows for the retrieval of predefined nudges fromPredefinedNudes.jsonbased on the study design schedule and in cases where LLM generation fails.OnDeviceNudgeService.swiftmirrors the capabilities currently handled server-side byplanNudges.tsUserDataService.swifthandles miscellaneous user data operations forOnDeviceNudgeService.swift.📚 Documentation
For usage of
OnDeviceNudgeService(), seeLLMNudgesDemoView.swift.✅ Testing
OnDeviceNudgeService.swiftare inOnDeviceNudgeServiceTests.swift.LLMLocalDemoView.swift.LLMNudgesDemoView.swift.DebugForm.swift.Code of Conduct & Contributing Guidelines
By creating and submitting this pull request, you agree to follow our Code of Conduct and Contributing Guidelines: