542be32bed
cuBLAS requires M/N/K dimensions to be multiples of 8 for BF16 tensor core dispatch. state_dim=51 (with OFI) caused silent fallback to scalar FMA ops, leaving tensor cores at 0.1% utilization despite BF16 enabled. Pad state_dim to next 8-multiple (51→56, 43→48) at network construction and zero-pad input tensors in forward pass across all DQN network types: Sequential, DuelingQNetwork, DistributionalDuelingQNetwork, NetworkLayers. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ml
10-model ML ensemble for the Foxhunt HFT system, built on Candle v0.9.1.
Models
- DQN (Rainbow) — deep Q-network with prioritized replay, dueling heads, noisy nets
- PPO — proximal policy optimization with GAE, LSTM policies, clip-higher
- TFT — temporal fusion transformer for multi-horizon forecasting
- Mamba2 — state space model for sequence prediction
- Liquid Networks — biologically inspired networks for non-stationary data
- TLOB — transformer-based limit order book analysis
- KAN — Kolmogorov-Arnold networks
- xLSTM — extended LSTM architecture
- TGGN — temporal graph neural network
- Diffusion — diffusion-based generative model
Key Modules
ensemble— model ensemble coordination and confidence aggregationhyperopt— PSO-based hyperparameter optimization with per-model adapterstrainers— unified training loops (DQN, PPO, supervised)inference—InferenceAdaptertrait for predictioncheckpoint— model checkpointing and restorationevaluation— walk-forward evaluation pipeline
Usage
use ml::dqn::DQN;
use ml::ppo::PpoTrainer;