jgrusewski 042c38fed2 fix(dqn,ppo): verification swarm round 2 — gamma^n, checkpoint config, circuit breaker ...

- IQN + C51 Bellman bootstrap: gamma → gamma^n for n-step returns
- DQNConfig::from_safetensors_file(): reconstruct config from checkpoint
  metadata instead of hardcoding dims/Rainbow flags in enhanced_ml.rs
- PPO update_value_only(): add circuit breaker guard (consistency with update())
- PPO eval: tensor-core alignment on hidden dims (must match PpoTrainer)
- enhanced_ml.rs: checkpoint-driven config loading with fallback

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

2026-03-04 00:40:08 +01:00

..

baseline_common

fix(ml): wire spread_cost_bps into RL training — commission + spread = total tx cost

2026-03-01 22:10:43 +01:00

cuda_test.rs

chore: clean up examples, update ML binaries and risk tests

2026-02-27 01:33:18 +01:00

download_baseline.rs

chore: clean up examples, update ML binaries and risk tests

2026-02-27 01:33:18 +01:00

evaluate_baseline.rs

fix(dqn,ppo): verification swarm round 2 — gamma^n, checkpoint config, circuit breaker

2026-03-04 00:40:08 +01:00

evaluate_supervised.rs

fix(observability): dedicated OTLP runtime for sync training binaries

2026-03-02 18:54:10 +01:00

hyperopt_baseline_rl.rs

fix(dqn): production hardening — LR scheduler, urgency costs, F32 traps, atomic IO

2026-03-03 23:04:53 +01:00

hyperopt_baseline_supervised.rs

feat(hyperopt): record elapsed_seconds metric in all hyperopt binaries

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

train_baseline_rl.rs

fix(dqn): production hardening — LR scheduler, urgency costs, F32 traps, atomic IO

2026-03-03 23:04:53 +01:00

train_baseline_supervised.rs

feat(supervised): record learning rate and epoch duration metrics

2026-03-03 00:10:00 +01:00