Factory Docs
Start here for the retained factory contracts. They are runnable lab material,
not an active project queue; use them only after the fresh-experiment gate in
../NEXT.md is met.
posttrainllm’s retained product loop is:
target -> data -> post-training -> eval -> package -> report
Files
../README.md— golden path through state, attempts, reviewed products, roadmap, and learning.overview.md— what the factory is and what counts as proof.post-training-factory.md— how data, post-training, evals, performance, packaging, and public artifacts fit together.../attempt-ledger.md— all 76 final worked, failed, regressed, inconclusive, superseded, and rejected attempts.../external-products-reviewed.md— external products and techniques reviewed or adopted.case-study-template.md— public artifact report shape: baseline, failed attempts, slices, trace review, performance, and blockers.batch-posttraining.md— batch rollout, offline scoring, compact adapter update, eval, and decision loop.lora-geometry.md— adapter effective-update diagnostics for rank and module targeting.- Closed autocorrect foundation (training was rejected after the bounded bake-off):
autocorrect-foundation.md— contract, evaluator, simulator, and manifests;autocorrect-model-shortlist.md— the measured base bake-off and selection;autocorrect-adapter-recipe.md— the frozen LoRA recipe, the encoder-decoder training path, and its load-parity evidence. run-schema.md— local run directory contract.run-lifecycle.md— durable phase/revision state, metadata-only operator commands, advisory discovery, and recovery.report-card.md— portable before/after proof contract: measurement states, decision semantics, JSON + static report, publication gate.report-card-cohort.md— the published report-card cohort, documented absences, and the mapping gaps the review found.enforcement.md— native validation plus stricter publish-check requirements.eval-protocol.md— baseline, regression, and ship/reject rules.packaging.md— specialist package layout and lock metadata.reports.md— before/after report template.public-artifacts.md— public artifact registry, release states, and blockers.../techniques/README— method-vs-recipe registry. Use this before selecting a post-training run so “try DPO/RLVR/LoRA” becomes a concrete recipe with data, eval, slices, and stop rule.
Rule
Use existing primitives first. Add new tooling only when it directly improves data preparation, post-training, eval, packaging, or reporting for the current factory target.
Do not treat a method name as a plan. A run is ready only when it has a recipe: target, failure mode, data, reward or labels, eval gates, slice gates, and a ship/retry/reject threshold.