Pinned Loading

1. ml-experiment ml-experiment Public

   LoRa fine-tuning of Phi-2 and StarCoder2 on Apple Silicon using MLX

   Python

2. ml-experiment2 ml-experiment2 Public

   LoRA vs QLoRA-style fine-tuning of StarCoder2-3B on Apple Silicon with MLX: memory/quality tradeoffs, one variable per run

   Python

3. phi2-serving phi2-serving Public

   Local REST API serving Phi-2 behind FastAPI, with an optional MLX LoRA adapter. Native MLX on Apple Silicon or portable PyTorch CPU in Docker; synchronous and streaming generation.

   Python

4. ai-assistant-demo ai-assistant-demo Public

   Security-first Jarvis-style assistant: permission-gated tool calls with single-use JWTs, MCP tool fingerprinting, pgvector memory with RLS, and fully local voice.

   Python

5. gated-sre-agent gated-sre-agent Public

   AI SRE agent for k3s homelabs: observe k8s events, diagnose via LLM, act only through human-approved gates. Eval harness ships before the agent.

   Python 1

6. whiskersync whiskersync Public

   Voice-AI building blocks + a gated ElevenLabs rebooking agent. 6/6 live evals, p50 1.9s voice latency, 316 tests.

   Python