feat: GPU/TensorCore integration — TensorFlow backend, GPU-accelerated reasoning, training, and memory
- 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>
2026-04-28 05:05:50 +00:00
|
|
|
"""Tests for fusionagi.gpu.tensor_attention."""
|
|
|
|
|
|
|
|
|
|
import pytest
|
|
|
|
|
|
feat: complete all 19 tasks — liquid networks, quantum backend, embodiment, self-model, ASI rubric, plugin system, auth/rate-limit middleware, async adapters, CI/CD, Dockerfile, benchmarks, module boundary fix, TTS adapter, lifespan migration, OpenAPI docs, code cleanup
Items completed:
1. Merged PR #2 (starlette/httpx deps)
2. Fixed async race condition in multimodal_ui.py
3. Wired TTSAdapter (ElevenLabs, Azure) in API routes
4. Moved super_big_brain.py from core/ to reasoning/ (backward compat shim)
5. Added API authentication middleware (Bearer token via FUSIONAGI_API_KEY)
6. Added async adapter interface (acomplete/acomplete_structured)
7. Migrated FastAPI on_event to lifespan (fixes 20 deprecation warnings)
8. Liquid Neural Networks (continuous-time adaptive weights)
9. Quantum-AI Hybrid compute backend (simulator + optimization)
10. Embodied Intelligence / Robotics bridge (actuator + sensor protocols)
11. Consciousness Engineering (formal self-model with introspection)
12. ASI Scoring Rubric (C/A/L/N/R self-assessment harness)
13. GPU integration tests for TensorFlow backend
14. Multi-stage production Dockerfile
15. Gitea CI/CD pipeline (lint, test matrix, Docker build)
16. API rate limiting middleware (per-IP sliding window)
17. OpenAPI docs cleanup (auth + rate limiting descriptions)
18. Benchmarking suite (decomposition, multi-path, recomposition, e2e)
19. Plugin system (head registry for custom heads)
427 tests passing, 0 ruff errors, 0 mypy errors.
Co-Authored-By: Nakamoto, S <defi@defi-oracle.io>
2026-04-28 08:32:05 +00:00
|
|
|
from fusionagi.gpu.backend import get_backend, reset_backend
|
feat: GPU/TensorCore integration — TensorFlow backend, GPU-accelerated reasoning, training, and memory
- 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>
2026-04-28 05:05:50 +00:00
|
|
|
from fusionagi.gpu.tensor_attention import (
|
|
|
|
|
attention_consensus,
|
|
|
|
|
cross_claim_attention,
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@pytest.fixture(autouse=True)
|
|
|
|
|
def _use_numpy():
|
|
|
|
|
reset_backend()
|
|
|
|
|
get_backend(force="numpy")
|
|
|
|
|
yield
|
|
|
|
|
reset_backend()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class TestAttentionConsensus:
|
|
|
|
|
def test_empty(self):
|
|
|
|
|
result = attention_consensus([], "query")
|
|
|
|
|
assert result["head_scores"] == []
|
|
|
|
|
assert result["consensus_score"] == 0.0
|
|
|
|
|
|
|
|
|
|
def test_single_head(self):
|
|
|
|
|
result = attention_consensus(
|
|
|
|
|
[["the sky is blue"]],
|
|
|
|
|
"what color is the sky",
|
|
|
|
|
)
|
|
|
|
|
assert len(result["head_scores"]) == 1
|
|
|
|
|
assert isinstance(result["consensus_score"], float)
|
|
|
|
|
|
|
|
|
|
def test_multiple_heads(self):
|
|
|
|
|
result = attention_consensus(
|
|
|
|
|
[
|
|
|
|
|
["the sky is blue", "water is wet"],
|
|
|
|
|
["security is important"],
|
|
|
|
|
["cost should be minimized"],
|
|
|
|
|
],
|
|
|
|
|
"what should we do about the project",
|
|
|
|
|
)
|
|
|
|
|
assert len(result["head_scores"]) == 3
|
|
|
|
|
assert 0.0 <= result["consensus_score"] <= 1.0
|
|
|
|
|
|
|
|
|
|
def test_with_weights(self):
|
|
|
|
|
result = attention_consensus(
|
|
|
|
|
[["claim a"], ["claim b"]],
|
|
|
|
|
"query",
|
|
|
|
|
head_weights=[2.0, 0.5],
|
|
|
|
|
)
|
|
|
|
|
assert len(result["head_scores"]) == 2
|
|
|
|
|
|
|
|
|
|
def test_empty_claims(self):
|
|
|
|
|
result = attention_consensus(
|
|
|
|
|
[[], []],
|
|
|
|
|
"query",
|
|
|
|
|
)
|
|
|
|
|
assert len(result["head_scores"]) == 2
|
|
|
|
|
assert result["head_scores"] == [0.0, 0.0]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class TestCrossClaimAttention:
|
|
|
|
|
def test_empty(self):
|
|
|
|
|
result = cross_claim_attention([])
|
|
|
|
|
assert result["similarity_matrix"] == []
|
|
|
|
|
assert result["conflict_pairs"] == []
|
|
|
|
|
|
|
|
|
|
def test_single(self):
|
|
|
|
|
result = cross_claim_attention(["only one claim"])
|
|
|
|
|
assert result["similarity_matrix"] == []
|
|
|
|
|
|
|
|
|
|
def test_two_claims(self):
|
|
|
|
|
result = cross_claim_attention(["claim one", "claim two"])
|
|
|
|
|
assert len(result["similarity_matrix"]) == 2
|
|
|
|
|
assert len(result["similarity_matrix"][0]) == 2
|
|
|
|
|
|
|
|
|
|
def test_self_similarity_high(self):
|
|
|
|
|
result = cross_claim_attention(["same text", "same text"])
|
|
|
|
|
sim = result["similarity_matrix"]
|
|
|
|
|
assert sim[0][0] > 0.9
|
|
|
|
|
assert sim[1][1] > 0.9
|
|
|
|
|
|
|
|
|
|
def test_conflict_detection(self):
|
|
|
|
|
result = cross_claim_attention([
|
|
|
|
|
"the project is very safe and reliable",
|
|
|
|
|
"completely unrelated topic about food and cooking",
|
|
|
|
|
])
|
|
|
|
|
assert isinstance(result["conflict_pairs"], list)
|