jgrusewski 4892d475ec config(ml-alpha): three sweep YAMLs for deployability validation (X19) ...

Adds smoke, threshold-tuning, and deployability sweep configurations
per spec 2026-05-19-ml-alpha-v2-trunk-grows-and-deployability-design.md
§3.3, §3.4, §3.6.

  sweep_v2_smoke.yaml          1 cell × W0 (cost=0.25, latency=200, th=0.75)
                                Verifies CheckpointV2 loads + harness OK.
  sweep_v2_threshold_tuning.yaml 8 cells × W0 (threshold sweep at realistic anchor)
                                Single-process fxt-backtest sweep. Picks
                                the pre-registered production threshold.
  sweep_v2_deployability.yaml  560 cells × W1-W4 (cost × latency × threshold)
                                Argo-fanned out. Feeds emit_deployability_verdict.

__SHA__ placeholders resolved by scripts/argo-lob-sweep.sh --sha at
submission time.

Per spec §3.3, §3.4, §3.6 (X19).

2026-05-19 08:45:12 +02:00

..

alpha_compose_backtest_c51_realspread.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_compose_backtest_c51.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_compose_backtest_pruned.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_compose_backtest.json

feat(alpha): --train-threshold for backtest + Phase E.3 honest verdict

2026-05-15 18:28:25 +02:00

alpha_dqn_h600_smoke_c51_realspread.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_dqn_h600_smoke_c51.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_dqn_h600_smoke_e2.json

feat(alpha): wire stacker-threshold controller into smoke rollout-end

2026-05-15 17:35:35 +02:00

alpha_dqn_h600_smoke_pruned.json

feat(alpha): Phase E.3 follow-up — C51 distributional Q + Thompson + L1-L10 depth + falsifications

2026-05-15 20:43:57 +02:00

alpha_dqn_h600_smoke_temporal.json

feat(phase-e-4-a): T13 smoke validation PASSES gate 1

2026-05-15 21:19:12 +02:00

alpha_dqn_h600_smoke.json

feat(alpha): wire Phase 1d.3 stacker into smoke — H=600 VERDICT PASS

2026-05-15 16:53:16 +02:00

alpha_dqn_h6000_smoke.json

fix(alpha): MBP-10 parser full-levels copy + fit_poisson L2 regularization

2026-05-15 17:20:52 +02:00

alpha_fill_coeffs.json

fix(alpha): MBP-10 parser full-levels copy + fit_poisson L2 regularization

2026-05-15 17:20:52 +02:00

alpha_random_baseline.json

fix(alpha): MBP-10 parser full-levels copy + fit_poisson L2 regularization

2026-05-15 17:20:52 +02:00

inference.toml

refactor(ml): delete mixed_precision module — BF16 unconditional on CUDA

2026-03-16 16:11:48 +01:00

model_params.toml

refactor(ml): delete mixed_precision module — BF16 unconditional on CUDA

2026-03-16 16:11:48 +01:00

simulation.toml

…

sweep_decision_stride_example.yaml

feat(fxt-backtest): sweep subcommand + example grid YAML (C15)

2026-05-18 09:34:08 +02:00

sweep_v2_deployability.yaml

config(ml-alpha): three sweep YAMLs for deployability validation (X19)

2026-05-19 08:45:12 +02:00

sweep_v2_smoke.yaml

config(ml-alpha): three sweep YAMLs for deployability validation (X19)

2026-05-19 08:45:12 +02:00

sweep_v2_threshold_tuning.yaml

config(ml-alpha): three sweep YAMLs for deployability validation (X19)

2026-05-19 08:45:12 +02:00