speculative-decoding

SKILLWorkflowcommunauté
v0.0.0Orchestra-ResearchMITMis à jour il y a 2 moisSource →

Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use when optimizing inference speed (1.5-3.6× speedup), reducing latency for real-time applications, or deploying models with limited compute. Covers draft models, tree-based attention, Jac

Community-submitted skill. Not yet reviewed by the Forge team. Full prompt content may not be available.Request review →
12kÉtoiles du dépôt
1Clients
1Formats
il y a 2 moisDernière mise à jour
Skill
AuteurOrchestra-Research
Version0.0.0
LicenceMIT
CatégorieWorkflow
Formatsskill.md
PromptOuvrir (voir l’onglet Prompt)
Compatibilité
Claude✓ Pris en charge
Cursor
Copilot
ChatGPT
Gemini
À propos

Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use when optimizing inference speed (1.5-3.6× speedup), reducing latency for real-time applications, or deploying models with limited compute. Covers draft models, tree-based attention, Jacobi iteration, parallel token generation, and production deployment strategies.

Mots-clés
skillclaude