hanemay/smartsightCoach-archive

VISIONPricing:Input $0.4 / Cached $0.08 / Output $0.8Concurrent Unit Cost:1Model Size:5.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jun 30, 2026License:cc0-1.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The hanemay/smartsightCoach-archive model, developed by Niclas Bade, is a 5.1 billion parameter LoRA fine-tuned version of Google's Gemma-4-E2B-it-qat-q4_0-unquantized base model, specifically designed for on-device fitness and nutrition coaching. This model integrates text-based coaching with vision capabilities for tasks like body-fat analysis, meal photo interpretation, and nutrition label reading. Its primary use case is to power the SmartSight AI coach application on mobile devices, offering a comprehensive, server-free personal coaching experience.

Loading preview...

SmartSight Coach LoRA Fine-tunes

The hanemay/smartsightCoach-archive model, developed by Niclas Bade, consists of LoRA fine-tunes for the SmartSight AI coach. This 5.1 billion parameter model is built upon Google's gemma-4-E2B-it-qat-q4_0-unquantized base, optimized for on-device inference on Android/iOS. It provides a comprehensive fitness and nutrition coaching experience directly on a user's phone, eliminating the need for server round-trips or accounts.

Key Capabilities

  • Integrated Vision and Text Coaching: Combines natural language interaction with advanced image analysis.
  • Body Fat Analysis: Accurately estimates body fat from photographs, with continuous improvements in accuracy and reduced bias.
  • Nutrition Label Reading: Extracts and interprets nutritional information from food labels.
  • Meal Photo Interpretation: Processes meal photos for dietary guidance.
  • On-Device Operation: Runs entirely on mobile devices, ensuring privacy and offline functionality.
  • Continuous Improvement: Undergoes rigorous iterative fine-tuning, with detailed evaluation against held-out data for each version.

Good For

  • Mobile AI Applications: Ideal for developers building privacy-focused, on-device AI solutions for health and fitness.
  • Personalized Coaching: Provides tailored fitness and nutrition advice without cloud dependency.
  • Vision-Language Tasks: Excels in scenarios requiring both textual understanding and image-based analysis in a specialized domain.
  • Resource-Constrained Environments: Optimized for efficient performance on mobile hardware, utilizing int4 quantization.

Finetunes
1 models