jgrusewski 6f645df11f feat(rl): wire lobsim book data from GPU SoA — proper rewards in GPU loader path ...

Add apply_snapshot_from_device to RlLobBackend: DtoD copy from
perception SoA buffers into lobsim book arrays + book_update kernel.
No host roundtrip, no sync (stream-ordered).

step_with_lobsim_gpu now calls apply_snapshot_from_device with batch 0's
last-snapshot (K-1) book data. Produces non-zero dones, pnl, wr.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-05-26 21:29:05 +02:00

..

cuda

feat(rl): R2 — fee model + loader pair API

2026-05-23 09:48:38 +02:00

src

feat(rl): wire lobsim book data from GPU SoA — proper rewards in GPU loader path

2026-05-26 21:29:05 +02:00

tests

refactor(rl): FRD horizons + range_σ are ISV-driven, not literals

2026-05-24 19:23:39 +02:00

build.rs

feat(ml-backtesting): build.rs + book_update CUDA kernel + sim skeleton

2026-05-18 08:26:53 +02:00

Cargo.toml

test(ml-backtesting): Ring 2 invariant fuzz at N ∈ {1, 16, 256} (C10)

2026-05-18 09:00:49 +02:00