fa71f973a6
Some checks failed
- New fusionagi/gpu/ module with TensorBackend protocol abstraction - TensorFlowBackend: GPU-accelerated ops with TensorCore mixed-precision - NumPyBackend: CPU fallback (always available, no extra deps) - Auto-selects best available backend at runtime - GPU-accelerated operations: - Cosine similarity matrix (batched, XLA-compiled) - Multi-head attention for consensus scoring - Batch hypothesis scoring on GPU - Semantic similarity search (pairwise, nearest-neighbor, deduplication) - New TensorFlowAdapter (fusionagi/adapters/): - LLMAdapter for local TF/Keras model inference - TensorCore mixed-precision support - GPU-accelerated embedding synthesis fallback - Reasoning pipeline integration: - gpu_scoring.py: drop-in GPU replacement for multi_path scoring - Super Big Brain: use_gpu config flag, GPU scoring when available - Memory integration: - gpu_search.py: GPU-accelerated semantic search for SemanticGraphMemory - Self-improvement integration: - gpu_training.py: gradient-based heuristic weight optimization - Reflective memory training loop with loss tracking - Dependencies: gpu extra (tensorflow>=2.16, numpy>=1.26) - 64 new tests (276 total), all passing - Architecture spec: docs/gpu_tensorcore_integration.md Co-Authored-By: Nakamoto, S <defi@defi-oracle.io>
FusionAGI Documentation
This folder contains architecture, API, and feature documentation for FusionAGI.
Documentation Map
flowchart TB
subgraph getting_started [Getting Started]
README[README.md at project root]
end
subgraph architecture [Architecture & Design]
ARCH[architecture.md]
IAD[interface_architecture_diagram.md]
end
subgraph api [API & Integration]
API_SPEC[api_middleware_interface_spec.md]
OPENAI[openai_bridge.md]
end
subgraph features [Features]
INTERFACES[interfaces.md]
UI_UX[ui_ux_implementation.md]
MAA_ACT[maa_activation.md]
MAA_MAP[maa_compliance_mapping.md]
MULTI[multi_agent_acceleration.md]
end
README --> ARCH
README --> API_SPEC
README --> OPENAI
ARCH --> IAD
API_SPEC --> INTERFACES
INTERFACES --> UI_UX
MAA_ACT --> MAA_MAP
Quick Reference
| Document | Description |
|---|---|
| architecture.md | Core components, data flow, AGI stack, self-improvement subsystem. Includes Mermaid diagrams. |
| interface_architecture_diagram.md | Full system and interface layer diagrams (ASCII + reference). Admin panel, core, multi-modal UI, end-user devices. |
| api_middleware_interface_spec.md | Dvādaśa HTTP/WebSocket API, SafetyPipeline, AdminControlPanel & MultiModalUI spec, Voice/Conversation layer. |
| openai_bridge.md | Using FusionAGI as an OpenAI-compatible model in Cursor Composer. Setup, config, auth. |
| interfaces.md | Admin Control Panel and Multi-Modal UI usage. Voice library, conversation tuning, modalities. |
| ui_ux_implementation.md | Implementation summary for admin panel and multi-modal UI. File structure, testing, next steps. |
| maa_activation.md | Manufacturing Authority Add-On: activation steps, MPC lifecycle, deployment. |
| maa_compliance_mapping.md | High-level mapping of MAA to ASME, ISO, MIL-STD, DoD Digital Thread. |
| multi_agent_acceleration.md | Parallel execution, agent pool, delegation, supervisor, batch/async routing. |
Visual Elements
Documentation uses Mermaid diagrams where supported (GitHub, GitLab, many Markdown viewers). Key visuals:
- README.md: System-at-a-glance flowchart, task lifecycle sequence.
- architecture.md: Component overview, data flow (task lifecycle), self-improvement flow.
- api_middleware_interface_spec.md: API → pipeline → core → interfaces flowchart.
- openai_bridge.md: Request flow (Cursor → FusionAGI API → Dvādaśa → response).
- multi_agent_acceleration.md: Supervisor and executor pool architecture.
- maa_activation.md: MAA activation and MPC flow.
- interface_architecture_diagram.md: ASCII system/interface diagrams (terminal-friendly).
For coding standards and project conventions, see .cursor/rules in the project root.