622d7cd757
Convert forward paths of TGGN, Liquid CfC, and multi-timeframe LSTM encoder from Candle Linear/Module/ops to cuBLAS-backed GpuLinear and gpu_* element-wise ops from ml-supervised. Candle Tensor remains at the UnifiedTrainable trait boundary; VarMap/AdamW retained for autograd. - TGGN: forward uses GpuLinear + gpu_relu instead of candle_nn::Linear - Liquid CfC: RNN loop uses GpuLinear + gpu_sigmoid/gpu_tanh/gpu_mul - MultiTimeframeEncoder: LSTM gates use gpu_matmul + gpu_sigmoid/gpu_tanh - Checkpoints save GpuLinear weights as JSON instead of safetensors - Remove candle_nn::Module, candle_nn::Linear from adapter imports Co-Authored-By: Claude Opus 4.6 (1M context) <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;