Free alternatives to spark-training-gotchas
50 free AI Skills apps to use instead, best rated first.
Top picks
- B
batch-grill-me
Getbatch-grill-me is an agent skill from Matt Pocock's skills collection, listed on skills.sh. It interviews you relentlessly about a plan until you share an understanding. It maps decisions as a tree and asks every question it can answer now
- W
webapp-testing
GetNative Python Playwright scripts for testing local web applications with server lifecycle management. Includes with_server.py helper to automatically start and manage single or multiple servers before running automation scripts. Supports re
- C
content-research-writer
GetAssists in writing high-quality content by conducting research, adding citations, improving hooks, iterating on outlines, and providing real-time feedback on
- O
openhue
Getopenhue is an agent skill from the OpenClaw project, listed on skills.sh. It lets an agent control Philips Hue lights and scenes through a Hue Bridge with the OpenHue CLI: turn lights on and off, dim rooms, set scenes such as movie mode, an
- I
implement
Getimplement is an agent skill from Matt Pocock's skills collection, listed on skills.sh. It implements work from a spec or tickets: it fetches the ticket, works test-first at agreed seams, runs type checks and tests as it goes, reviews the re
- F
frontend-design
GetDistinctive, production-grade frontend interfaces that reject generic AI aesthetics through intentional design choices. Guides aesthetic direction selection (brutalist, maximalist, retro-futuristic, luxury, organic, etc.) before implementat
More AI Skills apps
About spark-training-gotchas
spark-training-gotchas
Preflight and diagnose the ten known failure modes
spark-training-gotchas is an agent skill from the wshobson/agents collection that preflights and diagnoses ten known failure modes for machine learning training on NVIDIA DGX Spark. The DGX Spark uses a GB10 Grace Blackwell chip with an SM121 GPU, 128 GB of unified memory and an aarch64 CPU, and the skill names its ten recurring failure modes G1 to G10, across launch, memory, thermals, bandwidth and precision. Each is numbered so it can be checked by its number, and the numbering is used by checks that run these tests. The skill is meant to be read before a long run rather than after several hours of it. It applies when a training run fails to start with an import error or a segfault that does not point at the real cause, when a run runs out of memory while the GPU tools still show free memory, when throughput falls partway through a run that began well, before any multi-hour or multi-epoch job on GB10, when two Sparks are wired together and a parallelism strategy is needed, and when choosing between FP8 and NVFP4 precision for a run on the platform. The checks help the agent find which failure is likely, and the person running the job decides on the fix and confirms it on a short run. The skill is installed with the skills CLI from the wshobson/agents repository. Its skills.sh page showed about 2.3K installs, with Gen Agent Trust Hub and Socket passing and Snyk showing a warning at the time of the scan.
- Price
- Free
- Made by
- wshobson
- Category
- AI Skills
- Website
- github.com/wshobson/agents/spark-training-gotchas