d071501e44
Replaces scalar GAMMA_EFF_INDEX=43 with 4 per-branch slots at 43-46 (DIR, MAG, ORD, URG). New per_branch_gamma_update_kernel.cu reads per-branch v-range + health and writes 4 slots. Deletes gamma_update_kernel.cu per feedback_no_partial_refactor.md. Per-branch base/max: direction [0.92, 0.99] (trend horizon), magnitude [0.88, 0.95], order [0.85, 0.93], urgency [0.80, 0.90] (execution horizon). Kernel diverges each branch's γ from uniform bootstrap based on its branch's Q-stats. ISV slots downstream of 43 shift +3: KELLY (44→47), CQL_ALPHA (45→48), PLAN_THRESHOLD (46→49), Q_P05_* (47→50..54), Q_P95_* (51→54..58), fingerprint (55-56 → 58-59). ISV_TOTAL_DIM grows 57 → 60. Layout fingerprint auto-recomputes from updated seed bytes. c51_loss_kernel reads per-branch γ from ISV[43+d] inside the branch loop; iql_compute_per_sample_support reads ISV[43+d] per-branch for half_w. kelly_cap_update_kernel write index updated 44→47. q_quantile_kernel slot bases updated 47/51 → 50/54. state_layout.cuh ISV_PLAN_THRESHOLD_IDX updated 46→49. CPU-side PerBranchGammaMonitor (gamma_monitor.rs) is read-only observer of all 4 slots; read() returns mean, diagnose() exposes individual γ values + spread + health. No CPU-side γ computation (spec §4.C.6 GPU-drives). Plan 2 Task 3. Spec §4.D.2 + §4.C.6. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Foxhunt
Production HFT trading system in Rust.
Architecture
The workspace contains 32 crates organized as follows:
Core Libraries (16)
| Crate | Purpose |
|---|---|
trading_engine |
Order processing, FIX 4.4, IB TWS, SIMD, RDTSC timing |
risk |
VaR, Kelly, circuit breakers, kill switches, compliance |
risk-data |
Risk data types and shared structures |
trading-data |
Trading data types |
ml |
DQN Rainbow, PPO, TFT, Mamba2, ensemble inference |
ml-data |
ML data types and feature definitions |
data |
Market data ingestion and storage |
backtesting |
Replay engine, strategy tester |
adaptive-strategy |
Ensemble execution, microstructure analysis |
common |
Shared types, resilience, error handling |
storage |
S3 and local model storage |
model_loader |
Model serialization and loading |
market-data |
Market data feed handlers |
database |
PostgreSQL access layer (SQLx) |
config |
Configuration management |
tli |
CLI commands and tooling |
Services (8)
| Service | Purpose |
|---|---|
backtesting_service |
gRPC backtesting service |
broker_gateway_service |
FIX routing, broker connectivity |
trading_service |
Core trading operations |
ml_training_service |
Model training orchestration |
data_acquisition_service |
Market data acquisition |
trading_agent_service |
Autonomous trading agents |
api_gateway |
gRPC API gateway with auth |
web-gateway |
Axum REST + WebSocket gateway |
Frontend
web-dashboard/ -- React 19 + TypeScript + Vite + TradingView charts.
Building
# Check compilation (no PostgreSQL required)
SQLX_OFFLINE=true cargo check --workspace
# Run tests for a specific crate
SQLX_OFFLINE=true cargo test -p <crate> --lib
# Clippy
SQLX_OFFLINE=true cargo clippy --workspace
ML Models
Four production model architectures on Candle v0.9.1 with CUDA:
- DQN Rainbow -- Deep Q-Network with prioritized replay, dueling heads, noisy nets
- PPO -- Proximal Policy Optimization with GAE and LSTM policies
- TFT -- Temporal Fusion Transformer for multi-horizon forecasting
- Mamba2 -- State space model for sequence prediction
Each model has a standalone trainer and a UnifiedTrainable adapter for the hyperopt pipeline.
Infrastructure
- Git: Gitea at
git.fxhnt.ai(Tailscale-only), Scaleway DEV1-S - Observability: OpenTelemetry OTLP (env
OTEL_EXPORTER_OTLP_ENDPOINT) - Database: PostgreSQL with SQLx offline mode for CI
License
Proprietary. All rights reserved.