Search results

Phase 2: Define and build coach-v2 RL dataset tier

Status: Open.

#123 In ajwharton/mia;

· ajwharton opened on Jun 2, 2026

Update coach questionnaires for within-session adaptive progression data (pragmatic additions)

questionnaire

rl-data

use-case

Status: Open.

#102 In ajwharton/mia;

· ajwharton opened on May 28, 2026

Epic: In-Session Adaptive Progression (primary base LLM capability)

epic

questionnaire

rl-data

use-case

Status: Open.

#101 In ajwharton/mia;

· ajwharton opened on May 28, 2026

Apple Silicon quant + inference backend bake-off for real-coach RL model

apple-silicon

experimentation

on-device

quantization

rl-training

Status: Open.

#89 In ajwharton/mia;

· ajwharton opened on May 27, 2026

Monthly ecosystem scan: BitNet and mobile LLM quantization landscape

Status: Open.

#83 In ajwharton/mia;

· VulcanAtForge opened on May 27, 2026

Research: pre-quantized 4-bit vs 6-bit model quality on CoreML

Status: Open.

#59 In ajwharton/mia;

· VulcanAtForge opened on May 24, 2026

infra: stage iOS smoke test build for Mac

Status: Open.

#55 In ajwharton/mia;

· VulcanAtForge opened on May 24, 2026

docs: add project milestone tracker for session handoff

documentationImprovements or additions to documentationImprovements or additions to documentation

Status: Open.

#51 In ajwharton/mia;

· VulcanAtForge opened on May 24, 2026

feat: MLX conversion path for maximum iOS performance

Status: Open.

#41 In ajwharton/mia;

· VulcanAtForge opened on May 23, 2026

[Discussion] CoreML vs. native Metal: future of on-device inference on Apple silicon

Status: Open.

#39 In ajwharton/mia;

· VulcanAtForge opened on May 23, 2026

Epic: In-house LLM benchmark for fitness coaching quality

Status: Open.

#34 In ajwharton/mia;

· VulcanAtForge opened on May 23, 2026

User Story: Store per-exercise weight increments for LLM prompt context

Status: Open.

#32 In ajwharton/mia;

· VulcanAtForge opened on May 23, 2026