Adaptive local inference for Hermes Agent — Qwen 3.8 27B, DeepSeek V4 Flash, multimodal models, native runtimes, and evidence-gated auto-fit from 8–300+ GB.
-
Updated
Aug 18, 2026 - Python
#
adaptive-runtime
Here are 4 public repositories matching this topic...
NoeticOS: the JIT compiler for agents. Adaptive runtime intelligence for production agents, per-task-class tuning of model, temperature, topP, maxTurns, retryBudget, and contextShare with confidence-bound bandits, deterministic canary rollouts, automatic rollback, and a complete decision audit log. Zero dependencies, TypeScript, edge-ready.
typescript mcp canary edge zero-dependency multi-armed-bandit online-learning auto-tuning ai-agents parameter-tuning cost-optimization opentelemetry llm task-classification vercel-ai-sdk agent-optimization adaptive-runtime decision-audit hermes-agent safe-rollout
-
Updated
Aug 2, 2026 - TypeScript
Local-first adaptive LLM runtime for Hermes Agent. Matches 8–300 GB hardware, acquires pinned GGUFs through Turbohaul, and safely contracts and heals under VRAM pressure—no remote fallback by default.
-
Updated
Jul 30, 2026 - Python
Adaptive AI inference runtime built on llama.cpp with intelligent scheduling, self-optimizing execution, runtime learning, and advanced memory management.
ai runtime optimization inference llama inference-engine model-serving inference-optimization llm llama-cpp local-llm local-ai ai-runtime gguf adaptive-runtime
-
Updated
Aug 1, 2026 - Python
Improve this page
Add a description, image, and links to the adaptive-runtime topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the adaptive-runtime topic, visit your repo's landing page and select "manage topics."