Codeseys/composer-replication-framework

Wave 21: adversarial-review fixes — all 9 verified findings closed

Wave 21: Stage-0 dataset pipeline — swesmith engine, rollout harness, gates, contract

Wave 21: deep-read critical review — 8 source clusters re-read, findings verified

Wave 20: fix SageMaker smoke — torch-2.7 DLC + drop vllm pin (the real conflict)

Wave 20: fix SageMaker smoke — pin transformers<4.53 (torch 2.6 float8 skew)

Wave 20: SageMaker GRPO smoke artifacts (F3) — runnable g5.2xlarge GSM8K

Wave 20: Tier-0 fidelity fixes — k1-in-reward KL + Composer-2 behavior rewards

Wave 20: close F4 s3:// gap — live-S3 DiLoCo allreduce smoke (AWS_SMOKE-gated)

Wave 20: 5-facet AWS-native architecture design (F1-F5)

R14: harden live-Docker tests against transient daemon contention

docs: commit the Wave-2/3 review + Phase-8 verification audit trail (JSON findings)

Phase 8: B4-final (last 2 living-doc stale counts) + final verification disposition

docs: file R13 (pre-existing serverless ruff B904 debt) as tracked backlog item

Wave 3 cleanup: close deferred-LOW review items R5/R6/R11

Wave 3: close the HIGH review findings (kill-switch wiring, HeldoutSplit, EKS entrypoint bug)

Wave 2: 4 new modules (kill-switch, EKS/SageMaker executors, DockerSandbox) + B4/B7 completion

Wave 1: fix 8 failing tests + unblock Docker E2E + dep/doc debt

Merge docs/refine-2026-06: documentation refinement pass

docs(reconcile): main-lags-master foot-gun is RESOLVED, not live

docs: add refine-2026-06 engagement summary

docs: main is canonical branch; retire main-lags-master foot-gun

docs(wave3): add OVERVIEW.md, index ADR-014, fold in adversarial-review fixes

docs(wave2): archive point-in-time wave reviews + dated review bundles

docs(wave1): correctness pass — channel-3 provenance, gap honesty, dead links

docs+seeds: file remaining-work issues (Seeds) + project-state doc

seeds: sync 2026-06-09

feat(trainer): policy-optimization objective MENU (ADR-014)

docs(adr-013): close acceptance-gate boxes 1-5; only user-gated spend remains

fix(phase-8): close all 5 cross-family final-verify findings + regression tests

feat(b4-gpu+b6): GPU train-proof on A10G + docker-gated substrate E2E test

feat(wave-b): ADR-013 LMA integration + B4 end-to-end SDPO-fires proof + doc refresh

feat(wave-a): close ADR-011 (SDPO alignment indices) + ADR-012 (review findings)

architect: ADR-011/012/013 + research (alignment-index fix, review-findings closure, LMA channel-ladder)

review+fix: cross-family adversarial ADR review (owed item) + remediation

docs: wave summary — Composer 2.5 data-gen + targeted-RL textual-feedback

feat(datagen): ADR-010 FeatureDeletionEnv synthetic-data subsystem; accepted

feat(hints): ADR-009 layered HintGenerator; accepted

feat(trainer): ADR-008 gate-3 live GRPO+SDPO smoke PASS; ADR-008 accepted

feat(trainer): ADR-008 Dr.GRPO config + SDPO strict-alignment guard

docs(adr): add ADR-008/009/010 (Dr.GRPO+SDPO, layered hints, FeatureDeletionEnv)

research: Composer 2.5 data-gen + targeted-textual-feedback deep-research wave

examples: add sdpo_real_trace_train_smoke — close the forward+backward+step link

Wave 21c: verify PRIME-RL adapter parity against upstream source (byte-for-byte)

Wave 21b: skip zero-signal SDPO on empty-recovery error turns + real-trace validation

Wave 21: close both Wave 20 debt items — chat-template alignment + structural is_error

Wave 20: ModalSpawnExecutor — finish the Modal-backed serverless executor

Wave 19: production-grade SDPO via ComposerDataCollator + adapter + collator fixes

Wave 18: 14 backlog items closed + 3-reviewer cross-family review

Wave 17: close all 5 audit FLAGs + SDPO context alignment + serverless re-exports

Wave 16: install ergonomics + gradient evidence + SDPO end-to-end example