jgrusewski c91995e41d feat(ml): add RawPrediction + predict_raw() default to ModelInferenceAdapter ...

Backward-compatible trait extension. Default predict_raw() wraps
predict() result with tensor: None. Adapters can override to return
raw GPU tensors for GPU-side ensemble aggregation.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-03-02 17:41:08 +01:00

..

asset_selection

…

backtesting

…

benchmark

fix(ml): update remaining PPO num_actions: 3 → 45 in hyperopt and benchmark

2026-03-02 15:41:23 +01:00

checkpoint

cleanup: delete 92 scattered .md files and .serena artifacts

2026-03-01 22:42:36 +01:00

common

feat(ml): wire PPO to 45 factored actions via FactoredAction

2026-03-02 15:41:23 +01:00

config

…

cuda_common

…

cuda_pipeline

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

data_loaders

…

data_pipeline

refactor: delete 22 dead examples, 4 CSVs, consolidate data to test_data/

2026-02-26 01:24:02 +01:00

data_validation

refactor(ml): rename real_data_loader to data_loader

2026-03-02 15:41:23 +01:00

deployment

fix: repair broken use statement in ml/deployment/endpoints.rs

2026-03-01 00:55:51 +01:00

diffusion

feat(ml): diffusion trainer GPU-accumulated loss + batched grad norm

2026-03-02 17:30:19 +01:00

dqn

refactor(ml): move FactoredAction to common/action.rs (canonical location)

2026-03-02 15:41:23 +01:00

ensemble

feat(ml): add RawPrediction + predict_raw() default to ModelInferenceAdapter

2026-03-02 17:41:08 +01:00

evaluation

…

explainability

fix(ml): make IG completeness axiom test deterministic

2026-03-02 15:41:23 +01:00

features

refactor(ml): rename real_data_loader to data_loader

2026-03-02 15:41:23 +01:00

flash_attention

…

gpu

feat(ml): GPU saturation backlog — PPO rollout sync, VRAM-aware dims, KAN GPU splines

2026-03-02 15:41:23 +01:00

gpu_benchmarks

…

hyperopt

fix(ml): update remaining PPO num_actions: 3 → 45 in hyperopt and benchmark

2026-03-02 15:41:23 +01:00

integration

…

kan

feat(ml): GPU saturation backlog — PPO rollout sync, VRAM-aware dims, KAN GPU splines

2026-03-02 15:41:23 +01:00

labeling

…

liquid

feat(ml): GPU saturation final — DQN PER deferral, PPO trajectory batching, Liquid/TGGN sync reduction, DoubleBuffer wiring

2026-03-02 15:41:23 +01:00

mamba

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

memory_optimization

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

metrics

…

microstructure

…

model_registry

…

observability

…

optimizers

feat(training): move metrics to common::metrics, fix all clippy errors

2026-03-01 15:56:43 +01:00

paper_trading

…

ppo

fix(ml): align PPO action masking to canonical exposure*9+order*3+urgency layout

2026-03-02 15:41:23 +01:00

regime

fix(stubs): implement 15 production stubs, fix routing, delete placeholders

2026-02-28 23:47:31 +01:00

registry

…

risk

…

safety

…

security

…

stress_testing

…

tests

…

tft

feat(ml): TFT batched gradient norm — N GPU syncs → 1

2026-03-02 17:30:10 +01:00

tgnn

feat(ml): GPU saturation final — DQN PER deferral, PPO trajectory batching, Liquid/TGGN sync reduction, DoubleBuffer wiring

2026-03-02 15:41:23 +01:00

tlob

fix(ml,ci): zero-dim guards on all 10 models, eliminate warnings, unblock CI parallelism

2026-02-26 17:38:43 +01:00

trainers

feat(ml): TLOB batched parameter norm — N GPU syncs → 1

2026-03-02 17:29:33 +01:00

training

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

transformers

…

types

…

universe

…

validation

feat(ml): wire PPO to 45 factored actions via FactoredAction

2026-03-02 15:41:23 +01:00

xlstm

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

.gitignore

…

batch_processing.rs

…

benchmarks.rs

…

bridge.rs

…

cuda_compat.rs

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

data_loader.rs

refactor(ml): rename real_data_loader to data_loader

2026-03-02 15:41:23 +01:00

error.rs

…

examples.rs

…

feature_cache.rs

…

gradient_accumulation.rs

fix(ml): implement actual gradient clipping in PPO (was warn-only)

2026-02-27 10:03:42 +01:00

gradient_utils.rs

…

inference_validator.rs

…

inference.rs

fix(ml): remove hidden fake feature importance from inference path

2026-03-02 15:41:23 +01:00

lib.rs

feat(ml): add integrated gradients for feature importance

2026-03-02 15:41:23 +01:00

model_factory.rs

…

model_loader_integration.rs

…

model_registry.rs

…

model.rs

…

performance.rs

fix(ml): ignore perf benchmark tests in debug builds

2026-02-26 18:09:01 +01:00

portfolio_transformer.rs

…

preprocessing.rs

…

regime_detection.rs

…

tensor_ops.rs

…

training_pipeline.rs

feat(ml): comprehensive GPU saturation audit — 58 fixes across all 10 models

2026-03-02 15:41:23 +01:00

training.rs

refactor: delete 4,800 lines of dead ML code and remove MLFeatureExtractor dependency

2026-03-01 18:29:43 +01:00

traits.rs

…

walk_forward.rs

…