Rene Fichtmueller
|
2ca77d0aee
|
feat: Phase 2F — Multi-Agent Integration (ADRs + Client Fallback + Tests)
- ADR-0001: Multi-Agent Coworking Architecture with LLM Gateway Orchestrator
- ADR-0002: Tier Assignment Strategy for Model Selection (cost-first escalation)
- ADR-0003: Confidence Gate Thresholds & Learning Cycle Intervals (6h/12h/24h cycles)
- ADR-0004: External Provider Fallback Chain Ordering (Cerebras → Groq → Mistral)
- Enhanced client SDK: Offline Ollama fallback, health checks, exponential backoff retry
- Integration tests: claude-code-integration.test.ts (14 test cases)
- PHASE_2F_DEPLOYMENT.md: Pre-deployment checklist, automated deploy, rollback plan
- Post-deployment verification procedures for health, client fallback, metrics
|
2026-04-19 21:39:44 +02:00 |
|
Rene Fichtmueller
|
c3ab87b167
|
feat: add fo-blog-v8 training pipeline (Qwen2.5-14B, SFT+DPO)
Full v8 training pipeline for the optical networking blog model:
- train_blog_v8.py: SFT (LoRA r=64, 5 epochs) + DPO (2 epochs) on Qwen2.5-14B-Instruct
Fixed for trl 1.2.x: SFTConfig instead of TrainingArguments, processing_class= instead
of tokenizer=, eval_strategy= instead of deprecated evaluation_strategy=
- consolidate_v8_dataset.py: weighted merge of all data sources (820 effective SFT / 235 DPO)
- crawl_v8_sources.py: APNIC/RIPE Labs/potaroo/Cloudflare crawler with balanced div extraction
- process_v6_blogs.py: converts 101 real v6 TIP blog outputs into SFT + DPO pairs
- label_v7_quality.py: Claude-judged quality labels → v8 quality DPO pairs
- parse_real_posts.py: parses blog.fichtmueller.org Ghost CMS HTML → gold SFT records
- run_v8_pipeline.sh: autopilot (consolidate → SFT → DPO → GGUF → Ollama)
- blog-v8-training.yaml: training config reference
Dataset breakdown: 19 real posts ×3 + 196 v7-gen + 28 v6blogs ×2 + 135 external ×1.5
|
2026-04-19 11:44:09 +02:00 |
|