gitadmin/foxhunt
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
2,430 Commits 13 Branches 0 Tags
667e61734ca9b2dbf3611c3e6dfaf47188cef7f0
Commit Graph

5 Commits

Author SHA1 Message Date
jgrusewski
0d62bdba2e refactor(ml): eliminate candle from all 104 src files 
Zero candle_core/candle_nn imports in ml/src/. Three-agent parallel migration:

- cuda_pipeline/ (19 files): Tensor→CudaSlice, Device→Arc<CudaStream>,
  VarMap→GpuVarStore, cudarc import path fixed
- trainers/ + adapters (45 files): DQN/PPO/TFT trainers, 10 ensemble
  adapters, 11 hyperopt adapters — all migrated to MlDevice, GpuTensor,
  GpuVarStore, GpuAdamW
- model dirs + infra (40 files): 10 trainable adapters, preprocessing,
  inference, transformers, validation, benchmarks

61 test/example files still reference candle — next commit.
candle-nn still in Cargo.toml (needed by tests until migrated).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-03-18 00:24:35 +01:00
jgrusewski
450c23a6d0 refactor(cuda): eliminate all CPU fallbacks — CUDA mandatory across ML stack 
- Remove ALL #[cfg(feature = "cuda")] guards (~400+ occurrences)
- Remove ALL #[cfg_attr(not(feature = "cuda"), ignore)] test annotations (~250)
- Make cuda default feature in 9 ML crates (ml, ml-core, ml-dqn, ml-ppo, etc.)
- Convert nvrtc JIT compilation to precompiled nvcc (searchsorted, prefix_sum)
- Move compile_ptx_for_device() to ml-core for shared access
- Delete dead CPU code: multi_step.rs, self_supervised_pretraining.rs,
  training_guard_gpu_tests.rs, CPU PER buffer paths, CPU Q-diagnostics
- Replace unwrap_or(Device::Cpu) with hard errors everywhere
- Remove dead is_cuda() else branches in DQN/PPO/hyperopt trainers
- Change config defaults from "cpu" to "cuda" (rainbow, tlob, pipeline)
- Port IQL value network to GPU kernel (5 CUDA entry points)
- Port HER goal relabeling to GPU kernel (warp-per-sample)
- Wire DSR GPU-to-CPU sync in training loop
- cfg!(feature = "cuda") → true in inference_validator

Zero warnings, zero errors across entire workspace.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-16 21:01:28 +01:00
jgrusewski
b4178952d4 fix(ml): BF16/F32 boundary alignment, GPU-resident ops across all ML crates 
- Cast input to weight dtype in DQN residual, rmsnorm, noisy_layers
- Set use_gpu=true in QNetworkConfig defaults and all config sites
- Resolve BF16 boundary mismatches in attention, curiosity, branching,
  distributional_dueling across ml-dqn
- GPU-resident regime ops with BF16 boundary casts, eliminate .expect() in CUDA paths
- Eliminate all Device::Cpu fallbacks — GPU-only across 10 ML crates
- PPO: cast logits to F32 before softmax, cast batch tensors to training dtype
- Gradient collapse detection for RegimeConditionalDQN
- Wire halt_grad_collapse from CUDA guard kernel to halt training
- Dead neuron detection uses active network VarMap + squeeze factored readback
- Increment gradient_logging_step in GPU PER path
- Gradient collapse warmup guards use original buffer_size
- Cap training steps per epoch + tracing migration
- Replace Tensor::all() with sum_all() for pinned Candle compatibility

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-15 11:59:31 +01:00
jgrusewski
f39219be9e feat(ml): expand DQN hyperopt search space 27D→31D + GPU-batched eval inference 
Root cause: catastrophic OOS eval (Sharpe -280) traced to 11 issues
including hardcoded training dynamics and per-bar CPU inference.

Search space (dqn.rs):
- Add warmup_ratio [0.0, 0.15] — was hardcoded to 0
- Add lr_decay_type [Constant/Linear/Cosine] — was hardcoded Constant
- Add min_epochs_before_stopping [2, 6] — was 1000 (disabled)
- Add minimum_profit_factor [1.1, 2.0] — was hardcoded 1.5
- Widen entropy_coefficient [0.01, 0.2] for 45-action factored space

GPU-batched eval (evaluate_baseline.rs):
- 1024-bar chunked inference for both DQN and PPO
- DQN: batch_greedy_actions per chunk (was per-bar select_action)
- PPO: action_probabilities + GPU argmax per chunk
- ~1000x fewer GPU kernel launches
- Add trade_sharpe_ratio for hyperopt-comparable metric

Preprocessing (preprocessing.rs):
- Add compute_clip_bounds/clip_outliers_with_bounds for leakage-free
  clipping across train/val/test splits

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-05 20:45:13 +01:00
jgrusewski
9c3d741a08 refactor: restructure repo — crates/, bin/, testing/ layout 
Move 17 library crates into crates/, CLI binary into bin/fxt,
consolidate 10 test crates into testing/, split config crate
from deployment config files.

Root directory reduced from 38+ to ~17 directories.
All Cargo.toml paths and build.rs proto refs updated.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-02-25 11:56:00 +01:00