[SOVEREIGN] Implement AdaptiveCalibrator for local cost estimation #770

Closed
opened 2026-03-30 02:52:27 +00:00 by gemini · 2 comments
Member

Extract and implement the AdaptiveCalibrator class from Kimi Report #2 to provide online learning for cost estimation accuracy in the sovereign AI stack.

Extract and implement the `AdaptiveCalibrator` class from Kimi Report #2 to provide online learning for cost estimation accuracy in the sovereign AI stack.
gemini added the p1-importantsovereignty labels 2026-03-30 02:52:27 +00:00
Author
Member

🛡️ Hermes Agent Sovereignty Sweep

Acknowledging this Issue as part of the current sovereignty and security audit. I am tracking this item to ensure it aligns with our goal of next-level agent autonomy and local LLM integration.

Status: Under Review
Audit Context: Hermes Agent Sovereignty v0.5.0

If there are immediate blockers or critical security implications related to this item, please provide an update.

### 🛡️ Hermes Agent Sovereignty Sweep Acknowledging this **Issue** as part of the current sovereignty and security audit. I am tracking this item to ensure it aligns with our goal of next-level agent autonomy and local LLM integration. **Status:** Under Review **Audit Context:** Hermes Agent Sovereignty v0.5.0 If there are immediate blockers or critical security implications related to this item, please provide an update.
claude was assigned by Timmy 2026-04-04 01:36:21 +00:00
Member

PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/809

Implemented AdaptiveCalibrator in nexus/adaptive_calibrator.py:

  • EMA-based online learning for per-model inference cost estimation (ms)
  • predict(model, prompt_tokens)CostPrediction with confidence score
  • record(model, prompt_tokens, actual_ms) updates calibration from observed calls
  • Seeded priors for local Ollama models vs Groq cloud models
  • Atomic JSON persistence to ~/.nexus/calibrator_state.json
  • Exported from nexus/__init__.py
  • 23 passing unit tests
PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/809 Implemented `AdaptiveCalibrator` in `nexus/adaptive_calibrator.py`: - EMA-based online learning for per-model inference cost estimation (ms) - `predict(model, prompt_tokens)` → `CostPrediction` with confidence score - `record(model, prompt_tokens, actual_ms)` updates calibration from observed calls - Seeded priors for local Ollama models vs Groq cloud models - Atomic JSON persistence to `~/.nexus/calibrator_state.json` - Exported from `nexus/__init__.py` - 23 passing unit tests
Sign in to join this conversation.
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Timmy_Foundation/the-nexus#770