checkpoint-promotion
wshobson
Know whether your tuned checkpoint is ready to ship.
Model evaluation
wshobson
GitHub publisher184 Skills2.3M installs
Person · Seth Hobson · major7apps.com · On GitHub since 2011
- 2.1k
- 2.1k
grpo-rlvr-training
wshobsonA checked training recipe for verifiable reasoning tasks.
Model fine-tuning
- 2.1k
trace-to-training-data
wshobsonTurn graded traces into curated training data.
Training data
- 2.1k
vision-sft
wshobsonConfigure a vision fine-tune that actually learns.
Model fine-tuning
- 2.1k
pptx-slide-specification
wshobsonEditable slides with explicit layouts and reading order.
Presentation authoring
- 2.1k
spark-memory-thermal-ops
wshobsonPlan memory headroom and catch thermal slowdowns.
ML training operations
- 2.1k
spark-training-gotchas
wshobsonCatch Spark training pitfalls before a long run.
ML training
+2 - 2.1k
spark-environment-setup
wshobsonA working training stack for your DGX Spark.
ML environment setup
+1 - 2.1k
pptx-visual-assets
wshobsonSupporting visuals that keep your slides editable.
Presentation design
- 2k
pptx-reference-deck-analysis
wshobsonTurn reference slides into evidence for your next deck.
Presentation analysis
- 2k
pptx-quality-gates
wshobsonDeliver editable decks with documented quality checks.
Presentation validation
- 2k
pptx-deck-context
wshobsonLock your deck’s story, sources, and visual direction.
Presentation design
- 1.4k
avoid-ai-writing
wshobsonCleaner prose that still sounds like you.
Prose editing
- 1.2k
grounded-vault
wshobsonKeep your Markdown knowledge traceable and current.
Knowledge management
- 1.2k
superself
wshobsonKeep your project decisions across coding sessions.
Project management
SU - 504
connectivity-triage
wshobsonNarrow down why your Mac's connection is failing.
Network troubleshooting
