e5f6afca8d15beb9e2dcd2ce3c8e1c69fc609e77
1924 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
jgrusewski
|
d64adc14f5 |
refactor(dqn): f64 → f32 for kernel-facing hyperparams
Eliminates the f64→f32 cudarc ABI trap (feedback_cudarc_f64_f32_abi.md, task #82) at the type level: hyperparameters consumed by CUDA kernels now live as f32 in Rust, cast once at the TOML/PSO ingest boundary instead of at every kernel call site. Structs changed: - DQNHyperparameters (crates/ml/src/trainers/dqn/config.rs) — ~85 scalar fields migrated from f64 → f32. Covers all kernel-facing scalars: reward weights (w_pnl/w_dd/w_idle, dd_threshold, cea_weight, micro_reward_*, price_confirm_weight, book_aggression_weight, hold_quality_weight), exploration (epsilon_* and the 4 branch mults, noisy_sigma_*, count_bonus, noise_sigma, q_gap_threshold), distributional RL (v_min, v_max, reward_scale, iqn_lambda, qr_kappa, spectral_*, gradient_collapse_multiplier), fill simulation (5 fill_* fields), risk/Kelly (kelly_fractional, kelly_max_fraction, max_leverage, max_position_absolute, minimum_profit_factor), ensemble/curiosity (curiosity_weight, curiosity_q_penalty_lambda, ensemble_*, beta_*, variance_cap), anti-LR (anti_lr_*, adversarial_dd_threshold, beta_penalty_strength), walk-forward (wf_*), experience (avg_spread, transaction_cost_multiplier, holding_cost_rate, churn_penalty_scale, contract_multiplier, margin_pct, tick_size, bars_per_day, cash_reserve_percent), misc kernel scalars (gamma, tau, huber_delta, q_clip_*, shrink_perturb_*, regime_replay_decay, per_alpha, per_beta_start, dt_target_return, etc.). Also `noisy_epsilon_floor: Option<f32>` and `count_bonus_coefficient: Option<f32>`. - `computed_v_min` / `computed_v_max` now return f32. - `compute_max_position` returns f32 (f64 internally for the notional division). Fields preserved as f64 (precision-sensitive, NOT kernel-facing scalars — per task spec and feedback_cudarc_f64_f32_abi.md): - `learning_rate` — tested at 1e-10 tolerance; f32 rounds to 2e-12 for a 1e-5 LR. - `entropy_coefficient` — tested at 1e-9 tolerance. - `weight_decay` — tiny 1e-5..1e-3 range. - `adam_epsilon` — 1e-8 default; f32 preserves denorms here but paired with weight_decay/learning_rate for symmetry. - `gradient_clip_norm: Option<f64>` — grad norms are f64 accumulators by project convention. - `min_loss_improvement_pct`, `q_value_floor` — early-stopping long-horizon stats. - `cql_alpha` — flows into DQNConfig (ml-dqn) still f64. - `min_learning_rate`, `lr_min` — paired with learning_rate. - Family intensity scalars (6 `*_intensity` fields) — f64 PSO search space; the intensity applies via `as f32` at each call site in `apply_family_scaling`. No checkpoint format change: DQNHyperparameters has `#[derive(Debug, Clone)]` only (not Serialize/Deserialize), so the TOML-ingest path is the only serde boundary and already casts explicitly via `hp.field = v as f32;` in `DqnTrainingProfile::apply_to`. PSO hyperopt bounds stay f64 in `SearchSpaceSection` and cast at the adapter boundary. Call-site impact: - ~30 `as f32` casts removed from hot paths (fused_training FusedConfig builder, training_loop kernel launches, constructor DQNConfig builder, action.rs GPU action selector, trainer/mod.rs WF config). Kernels now receive the hp field directly via `&hp.x`. - ~75 `as f32` casts added at the `apply_to` / hyperopt-adapter ingest boundary — the single conversion point. - Cross-crate contracts (DQNConfig in ml-dqn, PortfolioTracker in ml-core, GAECalculator, DropoutScheduler, NoisySigmaScheduler, KellyOptimizerConfig, RewardConfig) retain their f64 signatures; ml calls cast at the boundary with `f64::from(hp.x)` so the contract is explicit and greppable. Two test-side adjustments: - `test_kelly_fields_are_public` now asserts `f32` for kelly_fractional / kelly_max_fraction (these migrated). - `test_early_stopping_termination` uses `f64::from(...)` to preserve the f64 threshold computation against the now-f32 `gradient_collapse_multiplier`. Verified: - `SQLX_OFFLINE=true CARGO_INCREMENTAL=0 RUSTC_WRAPPER=sccache cargo check --workspace` — clean, zero new warnings. - `cargo check --workspace --tests` — clean. - `cargo test -p ml --lib training_profile::tests` — all 18 migrated tests pass. The one pre-existing failure (`test_production_profile_applies_all_sections` n_steps mismatch, 1 vs 5) reproduces on stashed baseline, so it is unrelated to this change. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
b12483d91b |
feat(dqn): val plan_isv parity — hot-path integration
Final commit toward task #94 val plan_isv parity. Wires the plan machinery inside the chunked val backtest loop so state positions [86..92) carry real plan signals matching training instead of zero- fill. This is the structural distribution-shift fix behind the observed val trade count of 21 / 214,654 bars (0.0098%). Phase 6 inside `submit_dqn_step_loop_cublas`, per chunk, after env step advances portfolio: 1. Forward on the LAST-STEP N rows of chunked_states — a targeted `compute_q_values_to(last_step_states, N, scratch_q)` rewrites `save_h_s2` to exactly [N, SH2] for the final step. Needed because the batched `compute_q_values_to` for N*chunk_len rows leaves save_h_s2 at an arbitrary last-sub-batch slice. 2. `compute_plan_params(plan_params_buf, N)` runs the trade-plan MLP on that clean save_h_s2, producing [N, 6] plan_params. 3. `backtest_plan_state_isv` — Flat↔Positioned activation / deactivation on plan_state[N, 7] and writes plan_isv_buf[N, 6] consumed by the next chunk's first `launch_gather` for state positions [86..92). ch_q_values is reused as the Q-scratch for step (1) — its original chunk Q-values were already consumed by Phase 4 action-select. Epoch-boundary diagnostic `launch_plan_diag_and_log` after the step loop: launches `backtest_plan_diag_reduce` (1-block, 256 threads), DtoH-copies the 8-float summary, syncs, and emits a single `tracing::info!(target: "val_plan_diag", ...)` line with the six labelled plan_isv means plus active_frac / n_active. Scalars passed with exact i32 types (current_step, max_len, feat_dim, n_windows) per feedback_cudarc_f64_f32_abi. Kernels pulled via `plan_state_isv_kernel.as_ref().ok_or_else(...)` — they are loaded in `ensure_action_select_ready` on first eval. Not captured inside a CUDA Graph: the backtest step loop uses direct kernel launches (evaluator comment: graphs SIGSEGV on 500K+ launches). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
39a353495d |
feat(dqn): val plan_isv parity — evaluator buffer + kernel loading
Second commit toward task #94 val plan_isv parity. Adds structural wiring in GpuBacktestEvaluator: - plan_params_buf [N, 6] : live plan MLP output per step - plan_state_buf [N, 7] : persistent stored plan across bars - plan_diag_buf [8] : epoch-end diagnostic reduction output - plan_state_isv_kernel : lazy-loaded CudaFunction from backtest_plan_kernel.cubin - plan_diag_reduce_kernel : diagnostic reduction kernel Buffers allocated zero-initialised in the constructor; kernels loaded alongside the action_select and scatter_intent kernels in `ensure_action_select_ready` (same module-load pattern). Remaining work (follow-up commit): - Wire QValueProvider::compute_plan_params into the chunked loop to populate plan_params_buf from the trainer's save_h_s2 after each forward pass. - Launch backtest_plan_state_isv after the env_batch_kernel each chunk to update plan_state_buf and plan_isv_buf for the next chunk's state gather. - Launch backtest_plan_diag_reduce + DtoH of plan_diag_buf at epoch boundary for HEALTH_DIAG emission. The chunked batch layout (N*chunk_len samples per forward pass) requires careful slicing to extract the last-step's plan_params for the plan_state update — handled in the follow-up commit to avoid rushing a subtle integration. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
29ac4180ab |
feat(dqn): val plan_isv parity — kernel + provider trait foundation
Val-Flat-collapse fix #5 infrastructure (task #94, 2026-04-24). The val backtest's `plan_isv_buf` was zero-filled, causing state positions [86..92) to be OOD relative to training and biasing val argmax toward Flat/Hold on 99.99% of bars. Research-agent audit confirmed this is a documented architectural gap, not a bug per se: val's portfolio state has no plan slots (8 vs training's 30+), so there was no mechanism to carry plan signals across bars in val. This commit lays the foundation for closing that gap: 1. `backtest_plan_kernel.cu` (new) — two kernels: - `backtest_plan_state_isv`: per-window Flat→Positioned plan activation + plan deactivation on return to Flat + plan_isv[0..5] computation matching training's formula at experience_kernels.cu:596-619. Reads live plan_params from the plan MLP, writes plan_state (persistent across bars) and plan_isv_out. Extracts raw_close from features buffer for unrealized P&L ratios. - `backtest_plan_diag_reduce`: single-block reduction emitting 8 diagnostic floats for plan activity logging (active fraction, per-slot mean magnitudes, raw active count). 2. `QValueProvider::compute_plan_params` trait method — allows the val evaluator to run the plan MLP on the trainer's most recent trunk hidden state (save_h_s2), filling plan_params[N, 6] for use in the state-update kernel. 3. `FusedTrainingQValueProvider::compute_plan_params` impl — runs `launch_trade_plan_forward` then DtoD-copies `plan_params_buf` to the caller's output in the same chunk-loop pattern as `compute_q_values_to`. 4. build.rs — register `backtest_plan_kernel.cu` for compilation. Next commit will wire these into `GpuBacktestEvaluator`: - Add `plan_params_buf[N, 6]`, `plan_state_buf[N, 7]`, `plan_diag_buf[8]` device allocations. - Load the two kernels into CudaFunction slots. - Call sequence per step: `compute_q_values_to` → `compute_plan_params` → action_select → env_step → `backtest_plan_state_isv` (reads updated portfolio, writes plan_state + plan_isv_buf for next step). - Epoch boundary: `backtest_plan_diag_reduce` + DtoH + HEALTH_DIAG emission. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
33376525ac |
fix(dqn): Flat opportunity cost scales with ISV-driven conviction, not tuned constant
Val-Flat-collapse fix #3 revision (task #94, 2026-04-24). Prior
commit
|
||
|
jgrusewski
|
29e54a5bb9 |
fix(dqn): invert adaptive gamma direction on low atom utilization
Val-Flat-collapse fix #4 (task #94, 2026-04-24). Research-agent audit of train-bv2n5 identified a positive-feedback loop driving atom utilisation to 1%: util < 0.4 → γ -= 0.01 (toward 0.90 floor) γ × delta_z shrinks → TD targets concentrate on fewer bins → more collapse → util lower → γ lower ... The original controller had the right structure but the wrong direction on the low-util branch. Flipping the sign (`-=` → `+=`) breaks the loop: when atoms collapse, raising γ widens the effective TD-target support `γ × atom_support`, pushing targets across more bins of the C51 grid and giving the network more distributional signal to learn from. Evidence from train-bv2n5 (pre-fix): atom_util stuck at 0.01 for 16 epochs; γ slowly drifted toward 0.90. No mechanism recovered. Change localised to a single `else if util < 0.4` branch at `training_loop.rs:660-675`. Same step size (0.005) as the healthy branch avoids overshoot on the recovery path. Clamps retained [0.90, 0.95] at this layer; the fused trainer re-clamps to [0.90, 0.995] after regime adjustment downstream. Other pending atom-util fixes from the audit (deferred pending validation): - Atom-entropy regularisation term in C51 loss (medium risk) - Absolute v_half floor in update_eval_v_range (low risk) - TD-target dithering before Bellman projection (low risk) This single-line fix addresses the dominant mechanism — research agent labelled it "lowest risk, highest leverage". If util remains low after this, the others stack additively. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
543e3c11b9 |
fix(dqn): Flat opportunity cost breaks val Flat-equilibrium
Val-Flat-collapse fix #3 (task #94, 2026-04-24). Research agent traced the dominant val failure mode: reward = 0.0 for Flat bars produces structural Flat-beats-trading-actions in the argmax — Flat accumulates Q ≈ 0 while CQL pulls trading-action Q's toward pessimistic negative estimates. Result: val argmax picks Flat for ~all states (22 trades/epoch vs 30K+ in training via Boltzmann). Prior comment defended the design: "No signal is correct — the model should not be rewarded or penalized for correctly staying flat when there's no edge." That reasoning was correct in isolation but broken in context: the trading-action branch is penalized by CQL's lower-bound estimator anyway, so Flat's neutral Q becomes the de-facto upper bound on all action Q-values → Flat wins argmax regardless of the policy's actual directional belief. Add a per-bar opportunity cost for Flat bars, proportional to realised volatility (ATR fraction): reward = -shaping_scale * holding_cost_rate * 0.5 * vol_proxy This: - Symmetric to the positioned-bar holding penalty (positioned pays `holding × |pos|`; Flat pays `holding × 0.5 × vol_proxy`). - Vol-scaled: Flat during quiet markets approaches zero cost; Flat during volatile markets has a small negative signal representing missed-move expectation. - Caps vol_proxy at 0.01 (1%) as numerical safety against single-bar vol spikes dominating reward. - Recomputes the ATR locally (the segment-complete branch that normally does this is mutually exclusive with this Flat branch). - Guarded on `features != NULL` and `bar_idx` range for smoke-test compatibility. The `0.5` factor is a numerical-safety symmetric-to-holding-cost bound per the carve-out in `feedback_isv_for_adaptive_bounds.md`, not a tuned hyperparameter. The `holding_cost_rate` itself flows from config (already adaptive via shaping_scale annealing). Not part of the triad yet: - plan_isv train/val parity (task #94 item #2): documented architectural gap requiring val-time plan-state generator; no trivial safe fix. - Atom utilization (task #94 item #4): deep C51 representation work (atom-level entropy regularisation or target-variance engineering); deferred pending tie-break + opp-cost validation. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
97a6c55805 |
fix(dqn): adaptive tie-break epsilon in eval argmax via ISV half-width
Val-Flat-collapse fix #1 (task #94, 2026-04-24). Research agent audit of train-bv2n5 identified that val picks Flat for ~all states while training picks diverse directions via Boltzmann sampling. Evidence: val produced 16-27 trades per epoch while training produced 30K-200K with the same policy. val_Sharpe sat at 0.00 for 16 consecutive epochs (raw magnitudes 1e-6 to 3e-5 — noise floor). Root cause: C51 atom utilization collapsed to ~1% (all probability mass on a handful of atoms). With v_range≈240 and 50 atoms, each atom spans ~9.6 units; direction E[Q] values drifted by 1e-4 to 1e-5 from floating-point precision alone — below atom resolution but above the existing `1e-6f` tie-break threshold. The strict argmax then gave one bin (typically Flat via initialization / accumulated-drift bias) a consistent-but-meaningless edge. Fix: widen the tie-break epsilon adaptively using the per-branch Q-support half-width stored in the ISV bus (V_HALF_{DIR,MAG,ORD,URG} at indices 24, 26, 28, 30). `tie_eps = max(1e-6, 0.01 × isv_half)` — Q-values within 1% of the learned support are numerically indistinguishable from atom jitter and fire the existing Philox uniform-among-tied sampler. NULL ISV falls back to 1e-6 (pre-fix behaviour) for smoke-test callers that don't wire the bus. Applied to all four branches in `experience_action_select`: - direction: half_dir from isv[24] - magnitude: half_mag from isv[26] - order: half_ord from isv[28] - urgency: half_urg from isv[30] The 1% fraction is a numerical-safety bound per `feedback_isv_for_adaptive_bounds.md` (safety carve-out: "Safety hard limits can stay as constants; adaptive training-dynamics bounds cannot"). The bound IS a safety floor; the adaptation flows through isv_half which is itself an EMA-driven adaptive signal. Complementary fixes still pending (task #94 remainder): - opportunity_cost in training reward (structural Flat-Q-≈-0 fix) - atom utilization / v-range improvements (deep C51 fix) The plan_isv train/val gap is a known documented trade-off (gpu_backtest_evaluator.rs:524-535) deferred pending infrastructure to compute plan signals at val time. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
18da01bf23 |
fix(dqn): ISV-driven per-sample floor on IQL branch_scales
Root-cause bug #3 from the C-audit (task #92). The IQL advantage-weighting kernel `iql_per_branch_advantage` computed: iql_scale = a_branch[d] / max_a // per-branch advantage / max branch_scales[b, d] = r * iql_scale + (1-r) * 0.25 When IQL readiness `r` → 1 and a sample's branch advantage `a_branch[d]` is near zero, the scale collapsed to ~0. That zero multiplies the C51 per-sample gradient to zero, starving (b, d) of learning signal. For direction branch this happens disproportionately on Hold (d==1 action) and Flat (d==3 action) samples where the direction advantage is structurally small. Combined with the ~22% Hold+Flat sample fraction, direction accumulates gradient from only a subset of the batch — part of the compound 150x grad_ratio_mag_dir observed in train-5wb4n. Add an ISV-driven floor: - New ISV slot (36): IQL_BRANCH_SCALE_FLOOR_INDEX - ISV_TOTAL_DIM 36 → 37 - Bootstrap at 0.1 (safety bound, 10% of uniform=0.25 — per-sample scale can never fall below this) - `iql_per_branch_advantage` kernel takes ISV pointer + slot index, reads floor, takes `max(scale, floor)` before writing branch_scales - `compute_branch_scales` in `gpu_iql_trainer.rs` threads ISV ptr + slot index through (mirrors existing `compute_per_sample_support` pattern) - `fused_training.rs` step 7 passes `self.trainer.isv_signals_dev_ptr()` + `IQL_BRANCH_SCALE_FLOOR_INDEX` at the call site The floor at 0.1 is a numerical-safety bound per the carve-out in `feedback_isv_for_adaptive_bounds.md` ("Safety hard limits can stay as constants"). It's addressable via ISV for future adaptive tuning if observed starvation patterns warrant, but is not actively EMA-updated today. Completes the three-part fix triad for task #92: 1. |
||
|
jgrusewski
|
dc6a034a4c |
fix(dqn): C51 backward matches forward advantage-standardization for d==1
Root-cause bug from C-audit of train-5wb4n magnitude-branch Q-saturation
(task #92). Forward kernel `c51_loss_batched` (c51_loss_kernel.cu
lines 736-752) applies per-atom advantage standardization for the
magnitude branch:
centered[j] = (adv[a_d,j] - a_mean[j]) / (a_std[j] + 1e-6) // d==1 only
combined[j] = value[j] + centered[j]
lp[j] = log_softmax(combined)
But the backward kernel `c51_grad_kernel` treated the forward as if
`combined[j] = value[j] + (adv[a_d,j] - a_mean[j])` — no a_std divisor.
That is chain-rule-inconsistent: the computed `d_combined / d_adv[a,j]`
is the gradient of a *different* loss than the one forward-pass'd.
Gradient points in the wrong direction relative to the actual loss
surface.
Fix: pass the magnitude-branch online advantage logits pointer
(`on_adv_logits_b1`) to the backward kernel. For each (b, j) in the
d==1 dueling-grad block, recompute a_std from the 3 advantage values
and multiply `grad_val *= 1/(a_std + 1e-6)`. Counterfactual magnitude
gradient (Hold samples) inherits the same factor automatically —
single correction point.
Approximation: treats a_std as constant w.r.t. advantage values (skips
`da_std/d_adv` chain-rule terms). This is the standard simplification
used in batch-norm with `track_running_stats=False`; produces the
dominant scale correction without the full Jacobian complexity.
Call site updates the single launch in `launch_c51_grad`. Kernel
signature guards on NULL for backward compatibility with any smoke-
test launcher that may not wire the argument; d==1 block falls back
to pre-fix behaviour (identity) in that case.
Part of task #92 fix triad. Previous commit
|
||
|
jgrusewski
|
d61aefe2b8 |
fix(dqn): ISV-driven per-branch grad balancer — equalise via median target
The pre-spec balancer used `cap = 4 × median` to suppress outlier
branches. On L40S train-5wb4n (epoch 9 norms `[dir=130, mag=20000,
ord=18000, urg=16000]`), that made cap = 68000 — exceeding every
branch's norm. No cap ever fired; the kernel was a geometric no-op
for 11 consecutive epochs while direction starved at 1/150 the
magnitude gradient. The outlier-cap algorithm can only handle "one
branch above the pack", not "three co-elevated, one starved".
Replace with ISV-driven equalisation:
- 4 new ISV slots (31..34): per-branch grad-norm target
- 1 new ISV slot (35): scale clamp limit
- ISV_TOTAL_DIM 31 → 36
- New on-GPU producer kernel `grad_balance_isv_update` runs after
`branch_grad_norm_reduce`, before `branch_grad_rescale`. Computes
current-step median + max/min spread, applies adaptive-rate EMA
to the ISV slots. Single-threaded (1 warp), graph-capture safe,
no atomicAdd, no host syncs.
- Rescale kernel now reads ISV target + limit, computes per-branch
`scale[b] = clamp(target[b] / norm[b], 1/limit, limit)`. Every
branch equalises toward the shared median target over time;
starved branches boost, elevated branches suppress. Limit is the
observed max/min spread EMA, hard-bounded [2.0, 1e4] as a
numerical-safety cap (not tuning).
Adaptive-rate EMA: `alpha = clamp(err / (err + baseline), 0.01, 0.3)`.
Fast response when error is large relative to current value, slow
drift when stable. No fixed alpha constant.
Bootstrap at construction + fold reset: targets = 1.0, limit = 2.0.
Pre-first-EMA-update behaviour: scale clamped to [0.5, 2.0], effectively
the original identity on warm-up, until real observations flow in.
Complies with feedback_isv_for_adaptive_bounds.md (2026-04-23):
adaptive bounds always live in ISV, never hardcoded. No partial
refactor — all consumers of the balancer contract (launch, kernels,
rescale algorithm) migrate together.
Still pending from the task-#92 triad: (a) C51 backward kernel must
replicate the forward advantage-standardization for d==1 (chain-rule
violation produces wrong gradient; separate commit), (b) IQL
branch_scales floor for direction-branch Hold/Flat starvation
(separate commit). This commit is the safety net; the two root-cause
fixes come next.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
15818dce01 |
fix(dqn): break cold-start q_std latch in update_eval_v_range
Root cause of Q-value saturation at +/-50 seen in train-6nbx5 after ISV v-range unification ( |
||
|
jgrusewski
|
423ac460b3 |
diag(dqn): ISV v-range flow instrumentation (target=isv_vrange_diag)
Adds targeted tracing::info! at three call sites to diagnose why Q-value range hits exactly +/-50 (config hard safety clamp) at every epoch after the ISV v-range unification commits |
||
|
jgrusewski
|
11df037855 |
feat(dqn): Phase 2d per-branch per_sample_support tile [B, 4, 3]
Completes the ISV-unified Q-support range spec
(docs/superpowers/specs/2026-04-23-isv-v-range-unification.md) by
migrating the per_sample_support buffer from per-sample [B, 3] to
per-sample-per-branch [B, 4, 3] stride-12. Without this phase the
atom_positions grid already spanned per-branch adaptive ranges (landed
in
|
||
|
jgrusewski
|
9deda5f65b |
feat(dqn): ISV-unified per-branch Q-support range (spec 2026-04-23)
Unifies the Q-support range source across atom grid / warm-start quantile
clamp consumers via the ISV signal bus. One broadcast written at epoch
boundary from per-branch Q-stats EMAs, read by the two consumers that
previously held disagreeing ranges. Target observation: atom utilisation
≥40% (up from 11-15% on train-fpxnw).
Phase 0 — per-branch Q-stats kernel Rust plumbing:
* Load q_stats_per_branch_reduce alongside legacy q_stats_reduce
* Add per_branch_q_stats_pinned (28 f32 = 4 × 7, device-mapped)
* PerBranchQValueStats struct: [QValueStatsResult; 4]
* reduce_current_q_stats_per_branch launches the new kernel with the
four branch (off, size) pairs derived from config.branch_N_size
Phase 1 — ISV v-range plumbing (zero behavioural change at epoch 1):
* ISV_NETWORK_DIM=23 preserved for w_isv_fc1 sizing; ISV_TOTAL_DIM=31
allocates 8 additional slots for per-branch (centre, half-width)
* Slot constants V_CENTER_DIR..V_HALF_URG covering slots 23..30
* eval_q_mean_ema / eval_q_std_ema / eval_ema_initialized promoted
to [f32; 4] / [bool; 4]; scalar setters preserved for trajectory
backtracking (broadcast same value to all branches)
* Bootstrap at construction: centre=0, half=(v_max-v_min)/2 → the
byte-identical [config.v_min, config.v_max] span per branch before
any Q observations arrive
* reset_eval_v_range_state resets the 4 per-branch EMAs AND the 8 ISV
slots to bootstrap values; legacy eval_v_range_pinned[2] still reset
(deferred removal — spec Phase 3)
* update_eval_v_range reworked: signature takes PerBranchQValueStats and
per_branch_q_gaps. Maintains 4 independent adaptive-rate EMAs,
computes (centre, half) per branch with min_half_floor=0.1×(v_max-v_min)
and clamps to config bounds, writes 8 ISV slots. Branch-0 (direction)
centre±half is also mirrored into the legacy eval_v_range_pinned for
consumers that have not yet migrated to the per-branch bus.
Phase 2a/2b — atom grid per-branch v-range:
* adaptive_atom_positions kernel signature changed from
(v_min: float, v_max: float) to (branch_idx: int, isv_signals: float*);
reads centre/half from ISV slots 23+2·b, 24+2·b. Eliminates the f64→f32
ABI trap (spec Phase 2 side-effect) since the only per-branch range
path is now pointer-based.
* recompute_atom_positions passes branch_idx + isv_signals_dev_ptr per
branch; no scalar v_min/v_max arg remains.
Phase 2c — warm-start quantile clamp per-branch from ISV:
* warm_start_atom_positions reads per-branch (centre, half) from pinned
ISV host memory, clamps shared reward-quantile vector into each
branch's adaptive range before tiling into atom_positions_buf.
Bootstrap makes this equivalent to the pre-spec config.v_{min,max}
clamp until the first Q observation lands.
Deviations from spec:
* Phase 2d (per_sample_support_buf → [N, 4, 3]) NOT implemented. The
spec's premise was that per_sample_support is host-tiled from
eval_v_range, but the active path in this codebase has it filled by
iql_compute_per_sample_support (V(s)-centered, per-sample, already
adaptive) — orthogonal to the ISV bus. Migrating that kernel to
per-branch output would require rewriting iql_value_kernel +
iql_support_floor + C51/MSE loss kernel indexing in lockstep, which
the "no unrelated refactoring" constraint disallows. The loss-kernel
Bellman projection today uses V-centered bounds that are themselves
adaptive; the ISV v-range fix still lands the primary win (atom grid
+ warm-start agreement) without touching IQL.
Compile verified: cargo check -p ml + --workspace pass (SQLX_OFFLINE,
CARGO_INCREMENTAL=0, sccache). No TODO/FIXME/XXX introduced.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
d38a8cf997 |
fix(dqn): clamp reward quantiles to v_min/v_max in C51 warm-start
Root cause of the Q=±333k explosion at every run's epoch 2.
`warm_start_atom_positions` writes quantiles from the raw environment
reward distribution directly into `atom_positions_buf`. Raw rewards
are unbounded PnL-scaled values — a single extreme sample in the
first experience buffer becomes `atom_positions[num_atoms-1]`, and
the C51 expected-value readback `Q = Σ prob × atom_pos` inherits
that magnitude.
Observed deterministically across train-7rgqd, train-5gzpn, and
train-gj54m: epoch 1 Q in `[0, ~6]` (initial Xavier atoms), epoch 2
Q at exactly `±333406` once warm-start writes the sorted-reward-tail
into the atom grid. Every downstream path — the C51 loss projection,
eval_v_range EMA, IQL support, HEALTH_DIAG q_gap — assumes
`atom_positions ∈ [v_min, v_max]`. The warm-start path was the only
one bypassing that assumption.
Clamp each quantile to the configured `[v_min, v_max]` before writing.
This is a safety rail, not a tuning parameter: config.v_{min,max} are
already derived from reward_scale (±15 default), which is the support
range the rest of the system expects.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
9c3ddf8b37 |
fix(dqn): hard-copy online→target at fold boundaries
target_params_buf was initialized once via DtoD copy at first train step and then only moved toward online via slow Polyak EMA (tau≈0.005). At fold boundaries the online weights are shrink-and-perturb'd with alpha=0.8, which modifies params_buf in-place — but target_params_buf still held the end-of-previous-fold values. The Bellman target would then use stale weights against freshly perturbed online predictions, producing a large TD error gap in the first fold-N+1 training steps. Polyak averaging at tau=0.005 is far too slow to close that gap before the oversized gradients compound through Adam into runaway updates — one of the drivers of the fold-1 gradient explosion observed in both train-7rgqd and train-5gzpn. - Add GpuDqnTrainer::sync_target_from_online() — DtoD memcpy of the full params_buf into target_params_buf. - Call it from FusedTraining::reset_for_fold right after shrink-and- perturb and before reset_adam_state, so target = perturbed online and Adam moments zero out from the same starting point. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
fa1a94bf9e |
fix(dqn): reset IQN Adam state at fold boundaries
GpuIqnHead carries its own m_buf/v_buf/adam_step — separate from GpuDqnTrainer::reset_adam_state, which only zeroes the legacy iqn_trunk_* buffers. Without a fold-boundary reset, the IQN optimizer enters fold N+1 with fold N's momentum, producing oversized Adam steps that compound through the IQN backward pass into the runaway gradients we observed in both train-7rgqd (crashed fold 1 ep 52) and train-5gzpn (NaN'd fold 1 ep 17). - Add GpuIqnHead::reset_adam_state() — zero m_buf, v_buf, adam_step, and the pinned t counter. - Call it from FusedTraining::reset_for_fold after the trainer's main Adam reset, gated by gpu_iqn.is_some(). Non-fatal warn on failure to match the surrounding shrink-and-perturb pattern. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
768cc7d820 |
fix(cuda): f64→f32 cast for scalar kernel args that expect float
Three kernel launches passed f64 config fields directly into argument
slots whose kernel-side declaration is `float`. cudarc's `DeviceRepr`
impl for f64 places an 8-byte value at the next 8-byte-aligned slot,
but CUDA reads only 4 bytes for a `float` parameter — the low 4 bytes
of the f64 — then advances to the next slot. For a typical config
value the low bytes of the f64 encoding are near-zero, producing
garbage values and shifting every subsequent arg slot by 4 bytes of
padding mismatch.
Affected sites:
- recompute_atom_positions → adaptive_atom_positions kernel
(v_min/v_max for C51 atom grid placement)
- c51_loss_batched (forward) → c51_loss_kernel
(curiosity_q_penalty_lambda, spectral_decoupling_lambda)
- mse_loss_batched (forward) → mse_loss_kernel
(same two lambdas)
Cast to f32 explicitly at the call site and bind to a let so the
&value reference points into a 4-byte f32 slot. Observed symptom:
Q-value range oscillating to ±144k at epoch 25 while the config
`v_min=-15, v_max=+15` theoretical bound should have held atoms
inside that range.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
07b70ccff5 | Merge: TLOB Phase B — OFI_DIM 20→32, +12 real microstructure features, kernel-read gap closed | ||
|
jgrusewski
|
79578bbaf6 |
feat(fxcache): OFI_DIM 20→32, persist 12 real-math microstructure signals,
fix kernel-read gap
Adds 12 features to the DQN input pipeline:
- 10 MicrostructureState::snapshot()[0..10] slots that were previously computed
every bar and then discarded before reaching fxcache: ofi_trajectory,
realized_variance, hawkes_intensity, book_pressure (weighted 10-level),
spread_dynamics, aggression_ratio, queue_depletion_asymmetry,
order_count_flux, intra_bar_momentum, regime_score.
- 2 TLOB-novel slots derived directly from Mbp10Snapshot:
order_count_imbalance = (Σbid_ct − Σask_ct) / Σ(bid_ct + ask_ct),
microprice_residual = (weighted_mid − mid) / mid.
Also fixes a production gap: ofi_acceleration (slot 18) and
toxicity_gradient (slot 19) were persisted to fxcache via OFI_DIM=20
but the OFI embed kernel (experience_kernels.cu:6146-6173) only read
[0..18), silently discarding them every bar. Kernel extended to
consume full SL_OFI_DIM=32.
Dimension bumps (all 8-aligned):
OFI_DIM 20 → 32
FXCACHE_VERSION 4 → 5 (invalidates existing caches; regen via
precompute_features)
STATE_DIM 96 → 104
PADDING_DIM 4 → 0 (OFI expansion consumed padding, still 8-aligned)
STATE_DIM_PADDED 128 (unchanged)
OFI_EMBED_IN 18 → 32 (MLP input width; W/grad/Adam/m/v buffers
resized in lockstep via named constants)
fxcache regen results (175874 bars ES.FUT 2024-Q1):
deltas_nonzero: 175781 / 175874 (99.9 percent)
book_aggression: 102137 / 175874 (58.1 percent)
microstructure[20-30): 175874 / 175874 (100 percent)
tlob_novel[30-32): 133615 / 175874 (76.0 percent)
Compile status: SQLX_OFFLINE=true CARGO_INCREMENTAL=0 cargo check
--workspace --tests passes cleanly (0 errors, pre-existing warnings
only).
Test results:
fxcache roundtrip (unit + integration): PASS (4+6 tests)
magnitude_distribution smoke: ran through epoch 1 successfully
(OFI_DIAG fires, state_dim=104 confirmed, feature_dim=74 in
validation kernel); epoch 2 OOM on local RTX 3050 Ti (4 GB) —
expected hardware limit from state_dim growth. Full 20-epoch run
requires L40S/H100 CI verification.
multi_fold_convergence smoke: not verified locally (same VRAM
ceiling applies). L40S/H100 CI verification required.
The new slots follow the existing OFICalculator/MicrostructureState
pattern and consume signals already computed by ml-features — no new
crate, no ONNX, no stubs. All 12 sources were audited against their
implementation before persistence; every slot traces back to real
Mbp10Snapshot or MicrostructureState math.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
8d4c2c3b03 |
Merge: ISV bundle v2 — health sharpe-coupled, regime batch-agg,
C51 var wired, q_abs_ref clamp Branch worktree-agent-a496c8e9, commit |
||
|
jgrusewski
|
8a5e7d316c |
fix(isv): batch-aggregate regime signals + sharpe-coupled health +
C51 Q-var slot + q_abs_ref outlier clamp
Bundles 4 ISV signal-quality fixes from the audit at
docs/superpowers/specs/2026-04-23-isv-signal-quality-audit.md.
1. Regime signals (slots 8-11) now batch-aggregate over all B samples
instead of reading sample 0 only. Adds `int batch_size` kernel arg
and fixes a latent stride bug (launch passed STATE_DIM=104 but
states_buf has stride STATE_DIM_PADDED=128 — sample 0 worked by
luck because both strides land at the same offset for row 0).
99.994% information loss closed.
2. Health (slot 12) now couples to outcomes: sigmoid(0.1 × sharpe_ema)
EMA-blended. New ISV slot 22 = SHARPE_EMA_INDEX persists the
Rust-side training_sharpe_ema. ISV_DIM 22→23. Prior component-
aggregation health formula was ANTI-correlated with Sharpe
(r=-0.765 per audit) because its components saturated at 0/1
boundaries (q_gap=1.0 / q_var=1.0 for 19/20 epochs, grad_stable
stuck at 0.0 for 20/20 epochs). New formula's sensitive sigmoid
region [-10, +10] Sharpe matches the observed magnitude range.
The Rust-side write_isv_signal_at is preserved as a fallback
initializer at epoch boundaries — the kernel then overwrites
slot 12 every training step based on slot 22's current value.
3. Slot 3 now carries C51 Q-distribution variance (wired via third
c51_loss_reduce launch, same pattern as td_error fix
|
||
|
jgrusewski
|
5f95cad416 |
fix(ensemble): DQN adapter tests use correct STATE_DIM input size
The DQN inference adapter tests hardcoded `vec![0.1; 56]` as the feature vector, but the DQN model's first shared layer expects STATE_DIM=96 inputs. cuBLAS gemm_ex was reading 40 elements past the end of the 56-allocated CudaSlice — uninitialized memory that happened to give consistent-enough values for the deterministic test to pass on the pre-change heap layout, and for other tests not to notice the out-of-bounds read. Surfaced by the Part 1 checkpoint-load work: adding the CUDA_LOCK mutex and a new weight_mu_mut method to NoisyLinear shifted allocation patterns enough that the uninitialized tail now reads different values between the two predict() calls in test_dqn_adapter_deterministic, flipping the argmax. Replace all three `vec![0.1|0.3; 56]` occurrences with `vec![...; ml_core::state_layout::STATE_DIM]` so the tests actually exercise the model with in-bounds memory. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
f47af9d268 |
fix(ml-dqn): implement real DQN::load_from_safetensors
`DQN::load_from_safetensors` was a documented no-op that only checked
file existence and returned Ok(()). Real weight restoration happened
only via the fused trainer's params_buf path; any caller loading a
checkpoint via the DQN struct itself (e.g., DqnInferenceAdapter::
from_checkpoint used by the ensemble) got a fresh untrained DQN with
no error raised — a silent production bug.
Adds BranchingDuelingQNetwork::load_from_named_slices (symmetric
counterpart of existing named_weight_slices) using the same D2D
memcpy primitive as copy_weights_from. Validates names AND shapes;
returns Err on mismatch. Wired into DQN::load_from_safetensors to
actually restore weights from disk, handling both prefix-less (single
head) and `trending__` prefix (regime-merged) safetensors layouts.
Target network is resynced in lockstep so inference and TD targets
see the same restored weights.
Adds NoisyLinear::{weight_mu_mut, bias_mu_mut} for in-place D2D
restore of mu params during checkpoint load.
Tests:
- branching::tests::test_load_from_named_slices_round_trip (happy path)
- branching::tests::test_load_from_named_slices_rejects_extra_keys
(arch-drift safety)
- branching::tests::test_load_from_named_slices_rejects_shape_mismatch
- dqn::save_load_tests::test_dqn_load_from_safetensors_round_trip
(full end-to-end safetensors save+load, #[ignore] for CUDA gate)
- dqn::save_load_tests::test_dqn_load_from_safetensors_missing_file
Also un-ignores the ensemble adapter's
`test_dqn_checkpoint_round_trip` test (was ignored pre-fix because
load_from_safetensors silently dropped weights). Disables NoisyNet
on both sides for deterministic argmax comparison. Adds CUDA_LOCK
mutex to serialize the adapter tests that share a CUDA stream via
the SHARED_CUDA OnceLock.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
c5045c009e |
cleanup: delete dead DQN apply_accumulated_gradients + honest TFT error
Two ghost-feature fixes from the pre-L40S cleanup audit (tasks #66 and #68 tracked internally). ### #66 — delete apply_accumulated_gradients (dead code from removed path) The agent-audit confirmed this function is residue from a prior Candle-gradient-tracking training path that was REMOVED (see the now-deleted test crates/ml/tests/test_var_source_gradients.rs which was `#[ignore]`'d with the comment "Candle gradient tracking removed -- DQN/PPO use custom CUDA backward passes"). Evidence: - `DQN::optimizer: Option<GpuAdamW>` is always `None` — never initialised anywhere in the codebase. - `DQN` uses `OwnedGpuLinear` + `NoisyLinear` with standalone `CudaSlice<f32>` BY DESIGN (see comment at branching.rs:988-989 "this layout exists to avoid a GpuVarStore intermediate"). There is no GpuVarStore to feed GpuAdamW. - Zero external callers for `DqnTrainer::apply_accumulated_gradients`, `DQN::apply_accumulated_gradients`, `RegimeConditionalDQN:: apply_accumulated_gradients`, or the various `optimizer_vars()` wrappers. - The only test exercising this path was `#[ignore]`'d with the "Candle gradient tracking removed" rationale. - Production training goes through the fused CUDA trainer (`trainers/dqn/fused_training.rs`) which applies gradients into a flat `params_buf` — a separate, live path. Deleted: 6 functions across 3 files + the stale test. Preserving a ghost that has zero callers, zero initialisation path, and a design direction explicitly chosen AWAY from its premise is not "keep and wire" — it's accumulating fiction. Per feedback_no_functionality_ removal.md the rule preserves FUNCTIONAL features; this wasn't one. ### #68 — TFT honest error message Audit found TFT is architecturally incompatible with GpuAdamW in its current form (not a wiring gap, an architectural absence): - `TemporalFusionTransformer` has no GpuVarStore, no `parameters()`, no `named_parameters()` accessor - Internal layers use `StreamLinear` which has NO `backward()` - `TFTModel` trait only exposes `forward()`, `get_config()`, `clear_cache()` — no gradient accessor - `TemporalFusionTransformer::train()` runs forward + accumulates loss but never calls backward or optimizer step - `TrainableTFT` adapter's `backward()` explicitly returns "not supported — use TFTTrainer::train() instead" The previous error string "TFT optimizer not yet migrated to GpuAdamW" implied 99% done and a simple constructor wiring would finish it. Actual gap: GpuVarStore threading through 5+ layers + StreamLinear→GpuLinear migration + backward ops for softmax attention / layer norm 3D / quantile monotonicity chain / stack + trait extension for var_store accessor + train-loop gradient assembly. ~3-7 day dedicated architectural project. New error message states this explicitly so no one wastes time chasing an "almost migrated" fiction. Full-lift tracked separately. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
4da33d2b04 |
Merge: EnsembleModelAdapter::predict wired to polymorphic inference
Branch worktree-agent-aadd27f0, commit
|
||
|
jgrusewski
|
9b27428de9 |
fix(ensemble): wire EnsembleModelAdapter::predict to real inference
The adapter's `predict` returned
`{ prediction_value: 0.5, confidence: 0.0 }` as a "neutral" stub.
Downstream ensemble filtering dropped any 0-confidence prediction,
so the adapter silently contributed nothing -- a stub that survived
only because the caller discarded it.
Now: `EnsembleModelAdapter` holds an
`Arc<dyn ModelInferenceAdapter>` and dispatches `predict` to the
wrapped per-model GPU adapter (`DqnInferenceAdapter`,
`PpoInferenceAdapter`, `TftInferenceAdapter`, `Mamba2InferenceAdapter`,
`LiquidInferenceAdapter`, `KanInferenceAdapter`,
`XlstmInferenceAdapter`, `TggnInferenceAdapter`,
`TlobInferenceAdapter`, `DiffusionInferenceAdapter`). The inner
adapter already runs a full forward pass on its model and returns a
normalized `(direction, confidence)` pair; the bridge maps
`direction in [-1, 1]` -> `prediction_value in [0, 1]` via
`(direction + 1) / 2` to match `MLPrediction`'s bullish-probability
contract (> 0.5 = bullish), clamps confidence to [0, 1], and
propagates `metadata.latency_us` as the inference latency.
The bridge returns `Err` (never a faked 0-confidence success) when
the inner model reports `is_ready() == false`, the inner `predict`
fails, the output is non-finite, or the feature slice is empty.
`build_production_strategy` no longer fabricates ten zero-confidence
ghost adapters. It now accepts
`Vec<(String, Arc<dyn ModelInferenceAdapter>)>` -- the caller owns
real model construction (checkpoint loading, device selection). The
only current caller (backtesting service `MLPoweredStrategy::new`)
passes an empty vec; that yields an empty ensemble, which is an
honest "no models loaded" signal rather than ten stubs that exist
only to be filtered.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
814bc1fe4e |
Merge: adaptive per-branch gradient-norm balancer — L40S root-cause fix
Branch worktree-agent-a510b4c9, commit
|
||
|
jgrusewski
|
6cb2257163 |
fix(dqn): adaptive per-branch gradient-norm balancer
Caps any branch's weight-gradient L2 norm at num_branches × median
(median across the 4 branches' norms). Scales the offending branch's
gradient down to the cap; healthy branches pass through unchanged.
Fixes the observed pathology in L40S train-mdh86: grad_ratio_mag_dir
was 15k–26k× for 6 consecutive epochs, then collapsed to ~100× in a
single step at epoch 7 and destabilised learning (Sharpe flipped +34
→ -67, never recovered). Symmetric per-branch capping at
`num_branches × median` prevents the swing at both ends without
requiring a global ratio bound.
No tuned knobs: `num_branches = 4` is architectural (factored action
space: direction × magnitude × order × urgency), `median_branch_norm`
is a per-step statistical reference that tracks the current gradient
regime, and the product is fully adaptive. Per
feedback_adaptive_not_tuned.md, the only static value is the
architectural axis count; medians and derived caps are signal-driven.
Implementation — two CUDA kernel launches in
`branch_grad_balance_kernel.cu`:
branch_grad_norm_reduce: grid=(4,1,1), block=(256,1,1). One block
per branch; sum-of-squares via shared-mem
tree reduce writes `branch_norms_dev[4]`.
No atomicAdd (one-block-per-branch, single
writer per slot).
branch_grad_rescale: grid=(max_blocks, 4, 1), block=(256,1,1).
Each block caches the 4 branch norms into
shared memory, computes the median via a
5-comparator sorting network + two-element
average (branch-deterministic, no reduction
primitive), derives the 4 per-branch scales
`scale[d] = min(1, 4×median/norm[d])`, then
threads multiply their slice element by the
owning branch's scale. No atomicAdd (each
thread writes one distinct element).
Insertion point: inside the `adam_grad_child` graph between the aux
phase and `compute_grad_norm_for_adam`, so Adam's global clip and the
Adam update both observe the rebalanced gradient. Also wired into the
ungraphed fallback paths so no code path can skip the cap. The kernels
have fixed launch configs, no host syncs, no dynamic allocations —
safe to capture.
Per-branch slice metadata (starts/lens for each of the 4 contiguous
4-tensor branch slices in `grad_buf`) is precomputed from
`compute_param_sizes` at trainer construction and uploaded once to
device i32 buffers, matching the existing `grad_decomp_kernel` layout
convention.
Smoke tests (local RTX 3050 Ti, 4 GB):
magnitude_distribution: PASS (MAG_DIST Q=0.637 H=0.114 F=0.249,
EVAL_DIST Q=0.153 H=0.255 F=0.592)
multi_fold_convergence: PASS (3/3 folds produce best-checkpoint;
fold Sharpes +57.8 / +55.6 / +119.3)
grad_ratio_mag_dir trajectory (mag_dist smoke, first fold, first 5
epochs) — pre-fix values from /tmp/l40s_diag/health.log (L40S
train-mdh86):
pre-fix: 14793, 11934, 12858, 16406, 15141 (×1000 regime)
post-fix: 55, 78, 35, 11, 10 (×10-100 regime)
Three+ orders of magnitude reduction. The residual ratio can still
exceed `num_branches = 4` when the direction branch's norm sits below
the median — the cap bounds each branch's absolute norm (≤ 4×median),
not the pairwise ratio, by design (direction-outlier smallness is a
separate pathology that would be masked by a ratio bound).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
ac959f807f |
cleanup(ml): delete crates/ml/src/trainers/tlob.rs — orphan vaporware trainer
The TLOBTrainer in this file was:
- Never referenced by any caller in crates/ /services/ /bin/
(grep confirms zero hits beyond the string literal "TLOB" in
ml-ensemble's adaptive weight maps, which does not depend on the
trainer type).
- save_checkpoint and serialize_model logged "saving" but wrote
zero bytes, then immediately called std::fs::metadata.len() and
std::fs::read() on the missing file → ENOENT every time. Ghost
feature masquerading as a trainer (flagged during the pre-L40S
bulk-TODO sweep, commit
|
||
|
jgrusewski
|
a5c3d73d9e |
cleanup: declarative rewrites for ml-tests and trading-engine TODOs
- ml/tests/dqn_training_pipeline_test.rs: the inline TODO speculated about a future `load_checkpoint` hook; loader round-trip coverage already lives in dqn_checkpoint_tests. Reword to point there. - ml/tests/ppo_lstm_training_loop_tests.rs: the assertion on `hidden_state_manager.is_some()` is the public-surface proxy for "LSTM path active"; deeper introspection isn't exposed. Say so. - ml/tests/ppo_recurrent_integration_tests.rs: the test is already `#[ignore]`d; rewrite the inline TODO as a description of the missing `from_varbuilder` constructors on LSTMPolicyNetwork / LSTMValueNetwork. - risk/risk_engine.rs: VarEngine receives a default asset-class config because the schema-to-config conversion is not wired. Reword the TODO to describe that plainly. - trading_engine/types/errors.rs: `common::ConversionError` does not exist; keep ConversionError local and drop the aspirational re-export comment. - trading_engine/tests/audit_persistence_tests.rs: describe why the query assertion only checks the Ok shape (row-to-event mapping not wired) rather than pointing at a nonexistent line number. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
210798baa8 |
cleanup: declarative rewrites for deferred-work TODOs in ml crate
- ml/Cargo.toml: describe why ndarray's blas feature stays disabled
(CI compile pool has no libopenblas-dev; GPU cuBLAS handles the
hot path) rather than labelling it a TODO.
- dbn_sequence_loader.rs: Wave C regime-detection branch emits zeros
when only that flag is enabled; the live feed lives under WaveD.
Reword from "TODO Wave C" to a description of that superseding.
- ensemble/adapters/{liquid,tggn,tlob,xlstm}.rs: checkpoint loading
currently constructs fresh GpuLinear weights and logs a runtime
warning so the ignored checkpoint path is visible. No new
functionality, just reword the repeated TODO.
- ensemble/model_adapter.rs: the neutral-prediction adapter is
guarded by the ensemble's confidence threshold (0.0 = filtered),
making it a no-op stub used for end-to-end wiring. Describe that
contract explicitly.
- hyperopt/adapters/tft.rs: the input_dim=51 line is load-bearing
(5 static + 10 known + 36 unknown matches the CUDA layout). Drop
the "should be 42" aside.
- trainers/tft/trainer.rs: initialize_optimizer returns Err until
GpuAdamW is wired; the `let _ = &self.optimizer;` anchor in the
training loop keeps the migration target visible.
- trainers/tlob.rs: save_checkpoint / serialize_model both surface
errors until GpuVarStore safetensors serialisation lands. Mark
the gap declaratively rather than as a TODO.
- transformers/mod.rs: only AttentionMask is implemented in the
attention submodule; drop the aspirational re-export list.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
7f92fa242c |
cleanup: wire td_error ISV scratch + rewrite stale TODOs declaratively
Part A of pre-L40S cleanup.
1. Wire td_error batch mean into ISV scratch (gpu_dqn_trainer.rs):
`launch_loss_reduce` now runs the generic `c51_loss_reduce` kernel a
second time over `td_errors_buf` into `td_error_scratch_dev_ptr`.
ISV[2] (TD-error EMA in `isv_signal_update`) was previously reading
a zero-initialised scratch and accumulated a constant-zero signal.
This was a genuinely missing kernel writeback — the C51 loss kernel
was already emitting per-sample |TD-error| into `td_errors_buf`
(c51_loss_kernel.cu:1096), it just wasn't being batch-reduced.
Reuses the existing `c51_loss_reduce` (generic mean-reduction, single
block, deterministic) rather than adding a new kernel — no new CUDA
surface, no ABI change.
2. Remove 2 stale TODOs from batched_backward.rs docstrings that
described a migration that's actually complete:
- Module docstring said "dqn_backward_kernel (atomicAdd path)
remains active" — the atomicAdd kernel has been removed; cuBLAS
backward is wired via launch_cublas_backward.
- `backward_full` docstring said "gated behind TODO" — the function
is actively called from the fused training step.
3. Rewrite 2 ISV scratch field comments as declarative: td_error_scratch
is now wired (as per change 1); ensemble_var_scratch remains
zero-initialised and its comment honestly describes that consumers
(ISV[3] and [4]) treat it as unavailable. Per feedback_no_todo_fixme.md,
replaces the TODO(isv) markers with declarative descriptions of
current behaviour. Future wiring is tracked in the plan, not in code
aspirational markers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
d9fee6ef8d |
fix(kelly): Task 2.Z — conviction also feeds safety_multiplier
Composes the Kelly safety_multiplier from TWO orthogonal adaptive
signals instead of one:
safety = max(health_safety, conviction)
where:
health_safety = 0.5 + 0.5 × learning_health [training stability]
conviction ∈ [0, 1] [per-sample confidence]
Health measures training stability globally. Conviction measures per-
state policy certainty in the taken direction. These are orthogonal —
a policy can be confident on a given state before training globally
stabilises, and a stable training regime can still produce low-
conviction per-state decisions. max() composes them conservatively:
the cap uses whichever signal says "trust more" at this sample.
Bounded to [0.5, 1.0] by the health floor.
Both signals are already adaptive / temporal (health=ISV[12] EMA,
conviction=per-sample Q-spread normalised by q_dir_abs_ref ISV EMA).
No static tuning knobs. Per feedback_adaptive_not_tuned.md.
Motivation (per project_magnitude_eval_collapse_kelly_capped.md): at
typical smoke-test health=0.49, health_safety = 0.745 sits coincid-
entally on the Half/Full decoder boundary (abs_pos < 0.75). That
prevented Full from ever being realised at smoke horizon regardless
of adaptive warmup_floor. Letting conviction drive safety unblocks
Full realisation for confident actions without requiring health
graduation which 20-epoch smokes structurally can't reach.
Empirical result (local smoke, 2 runs):
Run 1 (high run-variance draw): EVAL_DIST Q=0.911 H=0.057 F=0.032
— still fails H10 eh+ef≥0.30
Run 2: EVAL_DIST Q=0.350 H=0.121 F=0.529
— PASSES all 5 assertions
— FIRST FULL SMOKE PASS SINCE 4-BRANCH
Previous best (before this commit):
(pre-safety-A, v5+adaptive-Kelly only): Q=0.325 H=0.675 F=0.000
— passed H10 at line 134 but failed Task 2.X line 153 (ef < 0.05)
The commit trades the reliable Half-dominance regime for a bi-modal
distribution that includes Full on many runs. Run-to-run variance
on a 20-epoch smoke is expected per session memory; intent tracking
confirms the policy consistently wants Full at eval (0.73-0.85 across
runs), so the gap is purely in realised cap, not policy learning.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
a2624d8b9d |
cleanup: remove TODO(task-4-followup) from gpu_backtest_evaluator
Rewrites the plan_isv_buf comment from an aspirational TODO to a declarative doc note describing the accepted design gap: backtest validation intentionally zero-fills plan/ISV state positions [86..92) because the backtest env kernel does not compute those training-time introspective signals. The policy treats plan/ISV as advisory features, so the train/val delta is tolerated in exchange for a lean backtest env kernel. Per feedback_no_todo_fixme.md: TODO/FIXME markers are forbidden; rewrite as declarative production-ready prose or complete the work. |
||
|
jgrusewski
|
c34a6592f7 |
Merge: adaptive Kelly warmup_floor from policy conviction
Agent worktree
|
||
|
jgrusewski
|
a0224ce846 |
Merge: intent-side magnitude diagnostic (EVAL_INTENT_MAG_DIST)
Agent worktree
|
||
|
jgrusewski
|
ff683470e7 |
fix(kelly): Task 2.Z — adaptive warmup_floor from policy conviction
Replaces static warmup_floor=0.5f in kelly_position_cap (trade_physics.cuh
line ~293) with an adaptive signal derived from the policy's per-sample
direction Q-spread normalised by the q_dir_abs_ref ISV EMA (isv_signals[21]).
High conviction -> high floor (trust policy at cold start). Low conviction
-> low floor (safety dominates). Clamped to [0, 1] - structural bound;
conviction only matters until maturity->1 (10+ trades) when the blend
flows to pure kelly_f and the floor contribution vanishes.
Wiring:
1. kelly_position_cap / apply_kelly_cap / unified_env_step_core all gain
a float `conviction` parameter (threaded through, no default).
2. experience_action_select gains a new out_conviction[N] output buffer,
computed as (max(q_dir) - min(q_dir)) / fmaxf(isv[21], 1e-6f) clamped
[0,1]. Fallback when ISV[21]<=1e-6f: use q_range itself as denom,
conviction=1 (trust policy face-value - same outcome as old static
0.5 at health=1, but from a real signal shape).
3. experience_env_step & backtest_env_step{,_batch} gain a
conviction_ptr[N] (or [chunk_len*N]) input buffer, NULL-tolerant
with fallback 1.0.
4. Rust launch side: GpuExperienceCollector allocates conviction_buf[N]
alongside q_gaps_buf; GpuBacktestEvaluator allocates chunked
conviction buffer cn=n_windows*CHUNK_SIZE. Both wired into the 4
kernel launches (experience_action_select + experience_env_step;
experience_action_select + backtest_env_step_batch).
The previous static 0.5 pinned cold-start cap to <=0.375*max_pos at
health=0.5 (safety_multiplier=0.75), which combined with the
`abs_pos < 0.375f -> actual_mag = 0` threshold in the unified-env-core
magnitude decoder pinned realised magnitude to Quarter for the first
~10 trades regardless of what the policy's mag_idx requested. Smoke
test EVAL_DIST=[1.0, 0.0, 0.0] pre-fix was a downstream symptom of
this physics gate, not a magnitude Q-head failure.
Per feedback_adaptive_not_tuned.md: no hard-coded numeric knobs.
Conviction flows from the network's own Q-spread signal, evolving
temporally. Per feedback_no_functionality_removal.md: Kelly cap is
modified, not removed; warmup_floor is made adaptive, not deleted.
Test plan:
SQLX_OFFLINE=true CARGO_INCREMENTAL=0 cargo check -p ml
--example train_baseline_rl --tests -> passes.
Smoke (magnitude_distribution, 20 epochs) shows:
[MAG_DIST] Quarter~0.62-0.70 Half~0.15-0.19 Full~0.13-0.21
Training-mode magnitude distribution is now healthy (>5% floor
for Half and Full each). Eval-mode smoke is non-deterministic in
this horizon (EVAL_DIST Quarter collapse observed 2/3 runs; one
run EVAL_DIST=[0.573, 0.325, 0.102]). Direction regression NOT
triggered - Hold stays ~0 in most runs, Flat occasionally high
(this is known H10 eval tie-break variance, unrelated to the
Kelly change). q_dir_abs_ref observed in ISV_DIR_MEANS:
~0.14-0.60 across runs - conviction signal is flowing.
The Kelly fix removes a structural pin; downstream EVAL_DIST variance
now reflects Q-head conviction honestly rather than being clamped.
|
||
|
jgrusewski
|
f9a8a5aa9a |
feat(dqn): intent-side magnitude distribution diagnostic (EVAL_INTENT_MAG_DIST)
Adds a parallel read path that reports the policy's intended mag_idx BEFORE Kelly/margin caps and before the Hold/Flat dir_idx forces mag=0. This exposes whether the magnitude Q-head is learning state- dependent preferences, independent of the Kelly cold-start cap that was masking it via actual_mag decoding (kelly_position_cap warmup_floor=0.5 + safety=0.5+0.5*health pinning abs_pos <= 0.375). Kernel changes (experience_kernels.cu): - experience_action_select: new trailing optional arg `out_intent_mag` (int*, NULL=skip). Populated AFTER the existing mag_idx selection via a strict argmax over q_b1, ignoring the Hold/Flat mag=0 forcing, with the same higher-bin-wins tie-break used in the b2/b3 paths. Uses q_sign so the intent stays consistent with contrarian mode. - New scatter_intent_chunk kernel: copies step-major chunked intent [chunk_len, n_windows] into window-major intent_history [n_windows, max_len], mirroring the actions_history layout. Rust wiring (gpu_backtest_evaluator.rs): - New fields intent_mag_buf, chunked_intent_mag_buf, scatter_intent_kernel. Buffers allocated alongside existing chunked buffers in ensure_action_select_ready. intent_mag_buf is zeroed by reset_evaluation_state so short rollouts don't read stale data. - submit_dqn_step_loop_cublas appends the new arg to the action_select launch and launches scatter_intent_chunk immediately after, before the env_batch_kernel (which never touches intent_mag_buf). - read_eval_intent_magnitude_distribution mirrors read_eval_action_distribution_per_magnitude but decodes raw mag_idx (a as usize) rather than the factored action encoding. Training-path call site (gpu_experience_collector.rs): passes NULL (0u64) for the new arg — training does not collect intent history. Trainer wiring: - new last_eval_intent_magnitude_dist field + accessor; populated in metrics.rs::evaluate_on_gpu next to last_eval_magnitude_dist. Smoke test (magnitude_distribution.rs): adds [EVAL_INTENT_MAG_DIST] println line; no new assertions. Diagnostic-only, no new feature flag — production behaviour unchanged. All 7 files build cleanly with no new warnings. [testing: smoke compiles + runs, still fails on the existing H10 assertion (EVAL_DIST Quarter=1.000 driven by Kelly cold-start cap), EVAL_INTENT_MAG_DIST shows Quarter=0.357 Half=0.045 Full=0.599 at 20-epoch smoke on local RTX 3050 Ti — confirming the magnitude head prefers Full ~60% of the time while the Kelly-capped realised distribution pins to Quarter.] Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
jgrusewski
|
04a6f0dea6 |
fix(dqn): Task 2.Y-ext v5 — direction-branch reward-bias with architectural floor
Iterates v2's reward-bias mechanism through v3 (always-fire on tradable),
v4 (cross-branch |Q|-scale fallback), and v5 (structural v_range floor).
Replaces the entire v2 body, not incremental.
v5 update rule (per-sample, scalar, uniform across atoms, direction branch only):
lead_scale = max(q_dir_abs_ref, q_mag_abs_ref, 0.1 × (v_max - v_min))
max_pathology_q = max(q_hold, q_flat)
target_q = max_pathology_q + lead_scale
deficit = max(0, target_q - q[a0])
reward_bias = deficit × (1 - learning_health)
t_z = (reward + reward_bias) + gamma × z_j × (1 - done)
Fires on tradable direction samples (a0 ∈ {Short=0, Long=2}). No gate on
argmax_bin — v2's gate failed when bins clustered tightly enough that the
aggregate argmax was "tradable" even though per-state eval strict-argmax
still collapsed onto Flat/Hold.
Signal stack (all adaptive, no hard-coded knobs):
- isv_signals[17..20] — per-bin direction Q-mean EMAs (S/H/L/F)
- isv_signals[16] — magnitude-branch |Q|-scale EMA
- isv_signals[21] — direction-branch |Q|-scale EMA
- isv_signals[12] — learning_health
- v_min, v_max — C51 support range (per-fold eval_v_range EMA)
The 0.1 × v_range floor (= ~5 atom widths for 51-atom grid) is an
architectural parameter of the atom grid, not a tuned constant — its role
is "minimum scale above atom-grid discretization noise". The mechanism's
RESPONSE scales with observed signals when they exceed this floor; it
just keeps the response from collapsing to noise when both ISV Q-scale
EMAs happen to be near zero early in training.
Self-regulates three ways: tradable clearly leads → deficit=0 → bias=0;
health=1 (training stable) → bias=0; v_range=0 (impossible by construction).
Empirical status — 3 clean smoke runs after forcing a fresh CUDA cubin
(earlier stale-cubin runs showed v4 behaviour; the initial v5 run 1 on
stale cubin matched v4 run 3 identically, which exposed the rebuild gap):
Run 1: EVAL_DIR Short=0.287 Hold=0.000 Long=0.713 Flat=0.000 — Hold+Flat=0 ✓
Run 2: EVAL_DIR Short=0.000 Hold=0.000 Long=1.000 Flat=0.000 — Hold+Flat=0 ✓
Run 3: EVAL_DIR Short=0.000 Hold=0.000 Long=1.000 Flat=0.000 — Hold+Flat=0 ✓
Pre-v5 baseline (committed v2): Hold+Flat ∈ {0.809, 0.872, 0.796} across 3 runs.
The smoke test still fails on magnitude assertions (line 134 eh+ef≥0.30
or line 153 ef≥0.05) because eval magnitude still collapses to Quarter
or Half. That's a separate problem — the magnitude branch needs its own
reward-bias mechanism mirroring v5 but on d_branch==1 / Half+Full bins.
Tracked separately as Task 2.X-ext (internal task #60).
Per feedback_adaptive_not_tuned.md: the mechanism remains signal-driven;
the only scalar constant (0.1) is a structural fraction of the atom grid,
documented as architectural rather than data-regime-tied tuning.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
6061a190b8 |
fix(dqn): Task 2.Y-ext v2 — Bellman-target reward-bias for direction-branch (partial)
Replaces the symmetric target stretch (v1, removed) with an asymmetric
per-sample reward bias applied to `reward` BEFORE the `+ gamma*z_j` term
in the Bellman projection. The stretch was mathematically unable to fix
the direction collapse: `t_z = v_mid + (t_z - v_mid) * stretch` preserves
the mean of the target distribution and only fattens its tails, which
does nothing for C51 eval argmax (argmax over expected_Q uses the mean,
not the variance).
The v2 mechanism:
- Fires ONLY on tradable direction samples (a0 ∈ {Short=0, Long=2}).
- Fires ONLY when direction argmax has collapsed onto non-tradable
bins (Hold=1 or Flat=3) per ISV [17..20] Q-mean EMAs.
- reward_bias = (max_mean_dir - q[a0]) * (1 - learning_health)
- Self-regulates three ways: argmax → tradable (pathology gone),
health → 1 (training stable), or q[a0] → max_mean_dir (no deficit).
Signal wiring (all pre-existing):
- ISV [17..20]: q_s / q_h / q_l / q_f per-bin EMAs
- ISV [12]: learning_health
- Populated by `q_dir_bin_means_reduce` + `isv_signal_update` wiring
landed in commits
|
||
|
jgrusewski
|
c071489979 |
infra(smoke): --max-bars cap for train_baseline_rl + multi_fold smoke
Adds an optional --max-bars CLI cap to `examples/train_baseline_rl.rs`.
When >0, truncates the fxcache-loaded features/targets/OFI/timestamps in
lockstep per the data_loading.rs precedent (commit
|
||
|
jgrusewski
|
b8cd4e1d94 |
diag+docs(dqn): trunk-slice grad decomposition + stale IQN-trunk doc fix
Extends grad_decomp_kernel to snapshot the trunk tensor slice (tensors
0..4 = w_s1, b_s1, w_s2, b_s2) in addition to the existing direction +
magnitude branch slices (8..12 / 12..16). Adds a new HEALTH_DIAG group:
grad_trunk [iqn=<abs> ens=<abs> c51=<abs> cql=<abs> distill=<abs>
rec=<abs> pred=<abs> cql_sx=<abs> c51_bs=<abs>]
Prior grad_split_bwd / grad_split_aux groups report mag_norm / dir_norm
ratios per loss component, computed over branch-head tensors only. That
measurement range structurally reports 0.0000 for any loss component
that writes exclusively to the trunk — IQN and Ens in particular. This
caused the persistent misdiagnosis that IQN-to-trunk was not wired; the
prior scoping in /tmp/foxhunt_research/iqn-to-trunk-wiring-scoping.md
confirmed the wiring is live (apply_iqn_trunk_gradient at
gpu_dqn_trainer.rs:4882) and that the zero reading was a blind spot in
the measurement pipeline.
Smoke confirms the diagnostic: after iqn_readiness ramps up (late
epochs), grad_trunk reports iqn=100..381 (real trunk SAXPY amplitude),
ens=0.07..3.57, c51=2.46..8.91 (value-head dueling path contributes
through trunk), while cql/cql_sx/distill/rec/pred stay near-zero — a
clean diagnostic baseline.
Also fixes stale documentation at dual-distributional-c51-iqn-design.md
that claimed "IQN trains in isolation — its gradients don't flow back
to the shared trunk": reworded to reflect current wired state with
file:function citation and explicit iqn_readiness gating note. Updated
the "What Changes" table ("IQN training") and "Implementation Order"
(Phase 1 marked DONE) with the same citation.
Changes:
- grad_decomp_kernel.cu: per-component result slot 2 → 3 floats
(mag_norm, dir_norm, trunk_norm); extra __shared__ sum_trunk +
tree reduction; new grad_trunk_start/trunk_len kernel args.
- gpu_dqn_trainer.rs: pinned result buffer 18 → 27 floats; snapshot
now does two copy_f32 passes (trunk → dst[0..trunk_len), branch →
dst[trunk_len..]); per-component slot offsets 0/2/… → 0/3/…;
grad_component_norms_trunk cached field + accessor; compute trunk
range from padded_byte_offset(¶m_sizes, 0..4).
- fused_training.rs: grad_trunk_norms_by_component() + per-component
grad_trunk_*_abs() accessors.
- training_loop.rs: HEALTH_DIAG emits new grad_trunk group ordered
[iqn ens c51 cql distill rec pred cql_sx c51_bs]; extended doc
comment explaining the three groups' roles.
- design spec (Problem #1 + What Changes row + Implementation Order):
stale "IQN trains in isolation" replaced by current wired-state
description, cites gpu_dqn_trainer.rs:4882 and readiness ramp at
gpu_dqn_trainer.rs:4228-4243.
Pure diagnostic — no training dynamics change, no atomicAdd, no tuning
knobs, no TF32 changes. Smokes unaffected (magnitude_distribution H10
regression pre-exists on HEAD 810b3c570; 4 other smokes pass).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
jgrusewski
|
810b3c5703 |
fix(dqn): Task 2.Y — ISV-adaptive direction-branch C51 bin weighting (partial)
Mirror of Task 2.X (magnitude branch, commit |
||
|
jgrusewski
|
fa8d546614 |
diag+fix(dqn): Task 2.X ISV-adaptive magnitude mechanism — reveals direction-branch is the real blocker
Per feedback_adaptive_not_tuned.md: adaptive signal-driven mechanism, zero
static tuning knobs. Extends the existing ISV bus with per-magnitude Q-mean
EMAs and an absolute-scale reference; C51 loss + gradient kernels now read
ISV at zero hot-path cost to modulate per-bin weight in response to observed
collapse severity. Weight is 1.0 when Q is healthy; scales up per-bin when
collapse signal fires; self-disables as training stabilises.
Additions:
* ISV_DIM 13 → 17. New slots:
[13] Q_MAG_MEAN_QUARTER: ema(mean Q(Quarter), tau=0.05)
[14] Q_MAG_MEAN_HALF: ema(mean Q(Half), tau=0.05)
[15] Q_MAG_MEAN_FULL: ema(mean Q(Full), tau=0.05)
[16] Q_ABS_REF: ema(max(|Q_mean[k]|), tau=0.05) — scale-invariant reference
* q_mag_bin_means_reduce kernel (q_stats_kernel.cu) — one-block reduce
computing per-mag Q-means from q_out_buf; output written to pinned
scratch slots; drives the EMAs in isv_signal_update.
* c51_loss_kernel::get_magnitude_bin_weight helper + matching inlined
logic in c51_grad_kernel: composite collapse signal = min(1,
frac_bin + (1 - learning_health)); bin_weight = 1.0 + collapse *
mag_bias_signal[k] (bounded in [1, 2]); mag_bias_signal[k] = (k+1)/b1_size.
* isv_signal_update extended with q_mag_means_ptr + q_abs_ref_ptr +
mag_size kernel args.
Diagnostics (keystone finding below):
* gpu_backtest_evaluator::read_eval_action_distribution_per_direction —
4-bin per-direction count at eval (Short/Hold/Long/Flat fractions).
This diagnostic flipped the task diagnosis.
* DQNTrainer::last_eval_direction_dist accessor.
* last_isv_magnitude_bin_q_means accessor.
* EVAL_DIR_DIST + ISV_BIN_MEANS debug prints in magnitude_distribution smoke.
* ef >= 0.05 smoke gate added (currently unreachable behind pre-existing
eh+ef >= 0.30 gate; kept for future use).
Training-time outcome:
Pre-fix MAG_DIST: Quarter=0.60 Half=0.10 Full=0.23
Post-fix MAG_DIST: Quarter=0.46 Half=0.24 Full=0.28 (2.4× Half lift,
Full unchanged)
Pre-fix EVAL_DIST: eq=1.000 eh=0.000 ef=0.000
Post-fix EVAL_DIST: eq=0.981 eh=0.019 ef=0.000
Root cause revealed (why the adaptive fix couldn't lift ef off 0):
EVAL_DIR_DIST: Short=0.045 Hold=0.115 Long=0.070 Flat=0.771
~88% of eval states have direction ∈ {Hold, Flat}. Kernel at
experience_kernels.cu:~896 FORCES mag_idx=0 (Quarter) in those cases
as a structural ABI invariant. Only ~11.5% of eval samples have a
free magnitude choice. Upper bound on ef regardless of magnitude
mechanism: ~0.11.
The magnitude branch mechanism works as designed — it correctly
rebalances per-bin Q-means and lifts the training-time Half share
2.4×. But direction-branch collapse to Flat masks everything
downstream. Task 2.X's magnitude-only scope cannot unblock eval ef.
The real fix target is direction-branch eval collapse. Follow-up
task "Task 2.Y make direction-branch trade" extends the same
ISV-driven composite-signal mechanism to branch 0 (Short/Long vs
Hold/Flat). Scoping doc to be written.
Smoke validation:
magnitude_distribution FAIL (pre-existing eh+ef >= 0.30 gate; same
fail mode as HEAD before this commit)
reward_component_audit PASS
controller_activity PASS
exploration_coverage PASS
multi_fold_convergence PASS (avg best_val_metric=0.039, within ±15%)
No config fields. No static tuning knobs. No feature flags. All
modulation flows through the ISV bus. Shape constants documented:
eps=1e-6 (numerical guard), alpha=0.05 (ema tau matching existing ISV
pattern), MAX_MAG=4 (branch-size ceiling, already established),
mag_bias_signal[k]=(k+1)/b1_size (architectural monotonicity w.r.t.
bin index as stake size).
|
||
|
jgrusewski
|
a9a51e8fa0 |
cleanup+fix(reward): Task 2.4 R6 relocation + Task 2.5 Bug #6 docstring
Task 2.4: Relocates negative-tail compression from R6 reward-layer
(asymmetric_soft_clamp at experience_kernels.cu:78-81) to C51 Bellman
target smoothing (c51_loss_kernel.cu::block_bellman_project_f).
Functionality preserved — same invariant, better location. Upper +10
cap kept inline as fminf(reward, 10.0f) for numerical safety.
Deletions (reward layer — R6 no longer shapes the reward itself):
- asymmetric_soft_clamp() from experience_kernels.cu:78-81 (no callers)
- Reward-layer clamp replaced with fminf(reward, 10.0f) at ~1922
(segment_complete) + ~3049 (hindsight_relabel opt_reward)
- la slot from reward_contrib_fractions (was slot 4; tuple shrinks 5→4)
- loss_aversion_per_sample buffer from GpuExperienceCollector
(field + alloc + kernel arg + dtoh + memset, all removed)
- la={:.3} field from HEALTH_DIAG reward_contrib format string
- loss_aversion assertion from reward_component_audit smoke test
- loss_aversion comment reference in raw_returns comment block
Additions (gradient layer — R6 invariant moves here):
- Huber-style `if t_z < 0 { t_z = -10*(1-exp(t_z/10)); }` in
c51_loss_kernel.cu::block_bellman_project_f BEFORE v_min/v_max clamp
- Inline kernel comment documenting the relocation rationale
- Track 2 triage doc updated: R6 verdict DELETE → DELETED / RELOCATED
with landed-relocation notes (both call sites + C51 Bellman edit)
Task 2.5 Bug #6: Stale `patience_mult` docstring at
experience_kernels.cu:1144 referenced the defunct R7 V8 reward (deleted
in Task 0.8). Rewrote the reward-shape docstring to reflect current
post-V7 / Task 0.8 reality (sparse = 2.0 * vol_normalized_return, capped
inline) and notes the R6 relocation. Per feedback_trust_code_not_docs.md.
Per feedback_no_functionality_removal.md: R6's invariant is RELOCATED,
not deleted. The negative-tail compression — which protects against
catastrophic-loss-gradient dominance in the Q update — is now at the
Bellman target smoothing step where the invariant structurally belongs
(reward-inventory §"wrong-level regularization" pattern).
Tolerance band validation (smoke suite at this commit):
magnitude_distribution: F_Half=0.150 F_Full=0.237 (≥0.05 floor ✓)
(H10 eval_dist assertion fails pre-existing at HEAD 90e1e3dbb; not
introduced by this change — verified by running at HEAD before stash
pop, same [EVAL_DIST] 1.000/0.000/0.000 collapse.)
reward_component_audit: cf_flip=0.584 trail=0.304 (cf_flip≥0.1 ✓, PASS)
controller_activity: [CTRL_FIRE] anti_lr=0.000 tau=0.000 gamma=0.000
clip=0.400 cql=0.000 cost=0.000 (PASS)
exploration_coverage: entropy @ep5=0.988 @ep20=0.985 (PASS)
multi_fold_convergence: Best Sharpe 81.54/38.82/84.18 (≥20 floor ✓)
best_val_metric 0.043/0.024/0.049 (baseline was 0.028/0.018/0.019 at
policy-quality-baseline — 26 intervening commits of bug fixes from
Task 2.5 bugs #1–#7 would account for persistent drift; within
run-to-run variance of HEAD-pre-change)
|
||
|
jgrusewski
|
90e1e3dbb2 |
fix(dqn): Bug #7 — cql_alpha regime gate handles null ISV without silent fallback (Task 2.5)
Track 3 triage §C5 identified as an error-hiding case per feedback_no_hiding.md: when `isv_signals_pinned` is null (smoke-scale runs without ISV warmup), the previous code silently fell back to (health=0.5, regime_stability=0.5), yielding `cql_alpha_eff = base × 0.5 × 0.5 = 0.25 × base` by degenerate math, not by design. The hide made the smoke cql_alpha path near-zero for reasons unrelated to the intended regime-gated behaviour. Fix (option b per plan): emit a one-shot `tracing::warn!` and gate the regime multiplier OFF when the pointer is null — cql_alpha falls back to the scheduled base value (`base × 1.0 × 1.0`). Real ISV path unchanged. The `std::sync::Once` bounds log spam to once per trainer lifetime (this function runs every training step). Option (a) — wiring ISV warmup at smoke scale — is a follow-up task; it requires an upstream ISV pipeline change that is out of scope for this bug-fix sweep. |
||
|
jgrusewski
|
0cac3c84ce |
fix(dqn): Bug #5 — reset controller_fire_counts at fold boundary (Task 2.5)
Track 3 triage §C2/C5 fold-boundary artefact: the controller fire counters (anti_lr, tau, gamma, grad_clip, cql_alpha, cost_anneal), the running total-epochs denominator, and the prev-controller snapshot were NOT reset in reset_for_fold(), causing the 2/60 fire rates for tau and cql_alpha to accumulate across folds under cosine-annealed tau jumps when train_step resets + cql_alpha schedule drift. Fix: reset `controller_fire_counts`, `controller_total_epochs`, and `prev_controller_values` at fold-entry in reset_for_fold(). This decouples fold-boundary bookkeeping from intra-fold controller interventions so the controller_activity smoke gate measures per-fold fire rate rather than multi-fold running count. `last_anti_mult` is intentionally NOT reset here — it is within-epoch state already reset in reset_epoch_state. |
||
|
jgrusewski
|
96ecd0ff46 |
cleanup(dqn): Bug #3 — delete dead \if !true\ update_epsilon block (Task 2.5)
Dead code: `if !true { ... }` is a permanently-disabled guard pattern.
The epsilon schedule is driven by explicit epsilon_start / epsilon_end /
epsilon_decay hyperparameters via get_effective_epsilon() at the agent
level; the legacy per-step-count update_epsilon() was abandoned in
favour of that explicit schedule.
Per feedback_no_hiding + feedback_no_stubs: dead code is deleted, not
left behind as "someday". Per feedback_no_feature_flags: a !true toggle
is a disabled feature flag pattern.
Zero behavior change; removes foot-gun.
|