Fine-tuning

Parent: LLMs.

Fine-tuning adapts base weights with domain data — LoRA/QLoRA and full finetune toolchains. Under LLMs it changes the model; it does not replace RAG, policy, or orchestration. Prefer RAG when knowledge changes weekly; prefer fine-tunes when format/style/domain language must live in weights.

Technically compare dataset pipelines, adapter merging, eval hooks, and license constraints on base models. Failure modes: finetunes that memorize PII, no frozen eval, and adapters nobody can reproduce from a commit.

Members: Unsloth, Axolotl, LlamaFactory. Prefer Unsloth for speed on limited GPUs; Axolotl/LLaMA-Factory when you need broader recipe control.

Features such as Learning & Adaptation and Bring Your Own Model mark the job. Always pair with Evaluations and then Serving to run what you tuned.

Inline members for audit: Unsloth Axolotl LlamaFactory. Features: Learning & Adaptation Bring Your Own Model Model Flexibility & Routing. Related terms referenced: Serving, Evaluations, LLMs.

For this research catalog the acceptance test is boring and strict: after wiring the pack, can a second person replay the same scenario and get the same measurable outcome? If not, the page is still a brochure. Document the scenario, the expected number, and the log line that proves it. That habit is what separates a deepened taxonomy cell from a landscape screenshot.

Also record what you refused: tools not connected, corpora not ingested, metrics not collected. Refusals keep the cell honest when marketing pages claim universality. A taxonomy essay that only lists winners teaches less than one that names the skip conditions next to the install path.

Keep the essay technical enough that an engineer can argue install order — and short enough that a reader finishes it before opening the right-rail member list. Prefer primary docs and repo evidence over landscape blurbs when the right-rail packs disagree.

Operational discipline for this cell: name the install artifact, the health check, and the rollback. Prefer one golden path documented against a member pack in the right rail over five half-wired demos. When packs disagree, believe the primary repository README and a failing test you can re-run — not the landscape one-liner. Measure latency, cost, and error rate on a frozen scenario before declaring the subcategory “covered.” If you cannot point at a log line that proves the scenario, the essay is unfinished even when the word count clears the audit floor.

Cross-link greedily but honestly: every taxonomy wikilink should change a reader’s next click, and every pack slug should resolve to a real pack. Do not invent members to decorate the prose. Prefer fewer true links over many decorative ones. Keep Monday-shaped typing out of the runtime story; observation and ledger remain senior to any agent UI that retypes reality.