spark-training-gotchas

SKILLWorkflowcommunity
v0.0.0wshobsonMITUpdated 11d agoSource →

Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.

Community-submitted skill. Not yet reviewed by the Forge team. Full prompt content may not be available.Request review →
39kRepo stars
1Clients
1Formats
11d agoLast update
Skill
Authorwshobson
Version0.0.0
LicenseMIT
CategoryWorkflow
Formatsskill.md
PromptOpen (see Prompt tab)
Compatibility
Claude✓ Supported
Cursor—
Copilot—
ChatGPT—
Gemini—
About

Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.

Keywords
skillclaude