training-pipeline-debugging SKILL.md skillA
training-pipeline-debugging from yeaight7/agent-powerups is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add yeaight7/agent-powerups/training-pipeline-debugging@git:20260605.2671f86Pins git:20260605.2671f86 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- yeaight7/agent-powerups
- Path
plugins/machine-learning-ops/skills/training-pipeline-debugging/SKILL.md- Versions recorded
- 3
- Latest version
- git:20260605.2671f86, committed Jun 5, 2026
- First version recorded
- May 2, 2026
- Size
- 2,404 bytes
- Repository stars
- 8
- License
- Apache-2.0
Use when an ML training run fails or misbehaves -- NaN losses, out-of-memory errors, shape mismatches, or a model that cannot fit even a single batch.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
## Purpose ML training bugs are often silent mathematical errors rather than explicit code crashes. This protocol localizes the standard failure classes quickly. ## When to Use - Loss becomes NaN or diverges - Training crashes with out-of-memory errors - Shape/dimension errors at matrix multiplications or loss computation - The pipeline runs but the model never learns ## Inputs - The failing training script and its error output or loss curve ## Workflow 1. **Run the overfit test first**: fit a single batch of ~10 examples. If the model cannot achieve near-zero loss on a single batch, the pipeline is fundamentally broken — do not debug full runs until the single-batch test passes. 2. **NaN losses** — check: - Learning rate too high? - Missing data (NaNs in input)? - Log/Exp/Divide-by-zero in custom loss functions? - Exploding gradients (clip gradients)? 3. **OOM (out of memory)**: - Reduce batch size. - Check for memory leaks in the training loop (e.g., accumulating history across epochs without `.detach()`). 4. **Shape mismatches**: - Add temporary print statements or assertions on `tensor.shape` before matrix multiplications and loss calculations. …
Read the whole file at its exact version.
How to install
mdr add yeaight7/agent-powerups/training-pipeline-debugging@sha256:accac340a4a8defbThe line at the top of this page pins the label git:20260605.2671f86, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_3mefdvr4utwlykhi)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 3 versions of this file, between May 2, 2026 and Jun 5, 2026, about one every 17 days. Its first recorded version is graded B and its latest A.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260605.2671f86 latest | 2026-06-05 | 2671f86 | 2,404 B | A | view · diff |
| git:20260502.647df10 | 2026-05-02 | 647df10 | 1,080 B | A | view · diff |
| git:20260502.317ce78 | 2026-05-02 | 317ce78 | 148 B | B | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (2404 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
yeaight7/agent-powerups · 8 stars · license Apache-2.0 · pushed 2026-09-27 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_3mefdvr4utwlykhi GET https://markdownregistry.com/api/v1/resolve?ref=yeaight7/agent-powerups/training-pipeline-debugging GET https://markdownregistry.com/api/v1/blob/accac340a4a8defb789352f92d05abf50adf94054fd59b8e3aeb975d62767bb1
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.