wassname cf9df71f6a add SKILL2.md: condensed anchor proposal (74 vs 703 lines)
Procedural/vibe anchor with gradual disclosure: calibrate + loop +
non-obvious numbers inline, tables/triage/sweeps demoted to on-demand
links into SKILL.md and refs/. Draft for side-by-side comparison; not
wired in (SKILL.md remains the entry point).

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
2026-06-02 13:26:59 +08:00
2026-04-09 05:09:25 +08:00
2026-04-09 05:09:25 +08:00

ML Debugging Folklore

Practitioner knowledge for debugging ML systems, curated and synthesized by wassname. Opinionated by source selection -- I picked sources I trust (Schulman, Goodfellow, CS231n, ...) and had an LLM extract the most relevant information for debugging ML systems.

Use as a Claude skill

/skills add https://github.com/wassname/ml_debug

Or paste SKILL.md into your system prompt / context when debugging.

What's here

  • SKILL.md -- the main artifact. Load into an LLM agent's context as a debugging skill. Parts 1-5 are reference knowledge; Part 6 is a runnable triage protocol (grep patterns, diagnostic snippets, decision tree); Part 7 is debugging mental models and practitioner priors.

  • docs/evidence/ -- frozen local copies of source material (blog posts, talks, papers, reddit threads). Claims in SKILL.md link back to exact quotes here.

S
Description
No description provided
Readme 1.5 MiB
Languages
Markdown 100%