Back to evaluations
Public evaluation
Health Recommendation Generation
The model will be evaluated based on its accuracy in predicting health outcomes, the relevance of its recommendations, and the user-friendliness of its interface.
Evaluation type
task based
Challenge
AI-Powered Fitbit Health Coach
Difficulty
Intermediate
Rigor
Unspecified
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
1
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
Health Recommendation Generation
Evaluates the model's ability to generate personalized health recommendations.
Input format
JSON containing Fitbit data (activity, sleep, etc.)
Output format
JSON containing personalized health recommendations and motivational messages.