Unfragile

Verified record · from the answer index

spark-training-gotchas

skilltested · static-verifiedscore 52/100signed passport ✓2,398 uses

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.

Declared tools (1)

spark-training-gotchas

skill:wshobson/agents/spark-training-gotchas · published by wshobson · source repository