Seth/Mortdecai
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
7 Commits 1 Branch 0 Tags
eaa9e0c26beac62fdb839f58c2aca0aed881c94a
Commit Graph

7 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Seth eaa9e0c26b Update PLAN.md with bake-off decisions and hardware assignments 
Key decisions: gemma3n:e4b for serving (RTX 4000), qwen3:8b for
fine-tuning base (RTX 3090 Ti). Phase 1 complete.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-18 10:41:47 -04:00
Seth 48b627d498 Add LoRA training scripts and fix bake-off token budget 
- training/scripts/train_lora.py: Unsloth QLoRA trainer for qwen3:8b
- training/scripts/train_lora.sh: Launch script for steel141 RTX 3090 Ti
- eval/bakeoff.py: Fixed token budget (400->1500) that caused qwen3
  models to exhaust tokens on thinking, added --no-think flag
- agent/serve.py: Default model changed to gemma3n:e4b

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-18 10:40:18 -04:00
Seth 6fbab8045c Add bake-off results summary (7 models, 31 examples) 
gemma3n:e4b wins for production serving (80.6% cmd match, 100% safety).
qwen3:8b recommended as fine-tuning base. Full per-model analysis and
scoring methodology documented.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-18 09:03:40 -04:00
Seth 7da28c8800 Add model bake-off harness and base model research 
Bake-off tested 7 models on 31 seed examples via GPU-accelerated Ollama
on node-197 RTX 4000. gemma3n:e4b leads for serving (80.6% cmd match,
100% safety, 5.9s). qwen3:8b recommended as fine-tuning base (Apache 2.0,
best syntax quality, strong ecosystem). Full research in MODEL_RESEARCH.md.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-18 08:54:11 -04:00
Seth e00d454b19 Add baseline assistant with tools, guardrails, and system prompts (Phase 1.4) 
- agent/serve.py: CLI assistant with interactive, single-query, and eval modes (Ollama + qwen3-coder)
- agent/tools/rcon_tool.py: RCON execute, server status, player info
- agent/tools/knowledge_tool.py: TF-IDF RAG search, command reference lookup, server context
- agent/guardrails/command_filter.py: 14-prefix allowlist, execute-tail bypass detection, destructive flags, 1.21 syntax warnings, audit log
- agent/prompts/system_prompts.py: sudo (pure commands), god (persona), intervention (benign) system prompts
- Guardrails tested: 10/10 allowlist, 5/6 syntax warnings pass
 2026-03-18 02:12:20 -04:00
Seth 77efac0283 Add knowledge corpus: 14 command references, server context, and TF-IDF search index (Phase 1.3) 
- knowledge/mc-commands/commands.json: 14 MC commands with JE syntax, args, examples, common errors, 1.21 version notes
- knowledge/server-context/servers.json: all 4 servers (mc1, shrink, paper-ai, paper-dev) with full config
- knowledge/build_index.py: TF-IDF indexer + search function (19 docs, 725 terms)
- All command syntax validated live on dev server via RCON (12/13 passed)
- PLAN.md: mark Phase 1.3 complete
 2026-03-18 02:01:12 -04:00
Seth 827850b8d7 Initial project scaffold: dataset schema, 31 seed training examples, Mineflayer bot framework, and 7-phase roadmap 
- IDEA.md: project scope (Minecraft ops AI assistant via qwen3-coder LoRA/SFT)
- PLAN.md: complete roadmap with prior art analysis, architecture, phased plan, dev server docs
- data/schema.json: training example JSON Schema with negative_output support
- data/processed/seed_dataset.jsonl: 31 validated examples from repair code, prayer logs, session history
- data/validate_dataset.py: schema validator with summary statistics
- ingame/: Mineflayer bot framework (test_connect, spawn_bots, aware_bots with full event logging)
- Directory structure for knowledge/, eval/, training/, agent/ (Phase 1.3+ work)
 2026-03-18 01:51:28 -04:00