Training Studio
Improve Fleet deliberately: diagnose first, train only the capability that is actually weak, and promote nothing until it beats its parent on held-out evaluation.
Checking champion, quality gates and training readiness.
General language, judgment, intent understanding, context isolation, tool choice and correction handling.
Server, application, coding and operational engineering.
AAA engine, world, rendering, animation, NPC and technical-art reasoning.
Python mastery engine Qwen3.8-27B
Verified micro-improvement loop. Unsloth/QLoRA is the preferred Qwen3.8 stack; DPO learns from executable chosen/rejected pairs. Nothing is promoted unless it beats the golden checkpoint without regression.
| Golden adapter | — |
|---|---|
| Model | Qwen/Qwen3.8-27B |
| Precision | 4-bit QLoRA |
| Protected passes | — |
| Latest self-improvement | — |
| Learning mode | — |
| Team learning | — |
Fleet Operator Bridge assistant-ready
One evidence-grounded control snapshot for you, Fleet AI, Forge and external assistants. It exposes the current target, source grounding, teacher panel, cartridge verification debt, mastery evidence, stall alerts and next action without making another AI reconstruct state from dozens of files.
| Current target | — |
|---|---|
| Evidence grounding | — |
| Progress strategy | — |
| Latest proof lane | — |
| Recommended next action | — |
Capability Autopilot 54-domain knowledge
Fleet's metacognitive control layer. It routes each task into the installed Supergenius domains, uses the smallest relevant knowledge slice, and diagnoses what is actually missing before asking for training.
| Operating rule | Know → retrieve → diagnose → acquire smallest missing capability → verify → continue |
|---|---|
| Self-diagnosis | Knowledge · skill · tool/code/runtime · orchestration · evaluation · model training |
| Mastery | Knowledge installation is not mastery. Exams, practicals, residency and retention stay separate. |
| Current status | — |
| Authoritative acquisition | idle |
Capability Fabric live App body
Fleet's dynamically discovered tool body. New Game Studio, Imaging, Reality, Audio, Browser, Coding and future capabilities become available to planning without model retraining.
| Operating rule | Discover capability → select smallest useful toolchain → act inside permissions → verify outcome → retain learning |
|---|---|
| Missing capability | Inspect existing modules/open source → prototype reversible App tool → verify → register automatically |
| Authority | Capability discovery never bypasses credentials, spend, GPU, production mutation, destructive or security gates. |
| Largest capability families | — |
Shared Causal World checking
One reusable Fleet actor/NPC runtime across RoutedLab and Flickbuster, with separate authoritative domain truth. Dialogue and 3D embodiment remain proposal/presentation layers only.
| Truth boundary | Authoritative domain state → bounded perception/belief → GOAP/procedures → validated actions → presentation-safe 3D projection |
|---|---|
| Memory | bounded · content-addressed · deduplicated · no Wasabi per-tick scratch |
| 3D contract | presentation only · hidden truth excluded |
| Training residency | causal-world-engineering · trained/ingested · fresh unseen exam queued · mastery not yet awarded |
Forge Engineering V42
Fleet's evidence-grounded network/architecture documentation and diagram compiler. Forge turns requirements, observed evidence and proof into consistent engineering deliverables without inventing missing facts.
| Lifecycle | — |
|---|---|
| Truth contract | — |
| Generic projection profiles | — |
| Source Authority Fabric | — |
| Structured vendor guidance | sentence-level constraints · source/hash receipts · version fail-closed · contradiction blocking |
| Known issues / field experience | — |
| Source drift | — |
| Vendor API adapters | — |
| Latest artifact QA | — |
| Integration debt | — |
Network Sensorium read-only evidence
Fleet's network-engineering instrument layer for Cisco IOS XE, FortiGate/FortiOS, Azure and generic network evidence. It extracts line-linked interfaces/routes/neighbors, keeps observed/inferred/unknown state separate, and plans the minimum evidence needed before a change.
| Safety contract | Read-only · no invented commands · no production mutation |
|---|---|
| Capabilities | — |
| Last snapshot | — |
Cognitive Reactor Matrix cartridges
Instant capability activation first. Fleet compiles verified knowledge into a runtime cartridge, plugs it into the skill/competence graph immediately, then Reality Gym proves transfer before any mastery claim or weight training.
| Mode | Matrix cartridge · instant runtime capability |
|---|---|
| Next stage | Autonomous verified practice → changed transfer → retention |
| Weight training | Only when a residual model-behavior gap remains after runtime skill + practice |
| Last event | — |
Training dataset
Built automatically from every teacher-escalated answer + verified skill. It grows as you teach Fleet AI.
GPU status —
The shared Vast controller. Fleet only ever wakes it for training — never automatically.
| State | — |
|---|---|
| Reserved by | — |
| Purpose | — |
| Hold remaining | — |
| Rate | — |
General knowledge LoRA workshop
Separate from the Qwen3.8 Python mastery engine above. This legacy/general workshop fine-tunes the current local assistant dataset and remains manual only.
Cognitive Clone Lab AI ZIP prototype
Capture transferable capability from a teacher AI into an installable Cognitive Clone Capsule: knowledge, causal models, procedures, failure patterns, decision rules, tool policies, counterexamples, preference pairs, tests and verification rules. This does not copy proprietary weights.
| Pipeline | interrogate → capture → compress → ZIP → Matrix install → changed-scenario proof → optional adapter distillation |
|---|---|
| Teacher sources | OpenAI · Anthropic Claude · Gemini · DeepSeek · open-weight teachers |
| Weight policy | Never claim proprietary provider weights were copied; only open/owned weights may be imported directly |
| Latest capsule | — |
| Balanced gene champions | — |
| Verified natural selection | No core-lane champions yet |
| Shadow Champion Lane | idle |
Matrix Cognitive Delta Compiler ready
Rapid capability import without replacing Fleet's Qwen3.8-27B core. Measure only missing competence, import structured deltas from OpenAI / Anthropic / Gemini / DeepSeek, preserve provenance, quarantine weak claims, compile an instant Matrix cartridge, then prove it in changed practice.
Recent training runs
No managed GPU training runs are active or recorded in this session.