fine-tuning-expert

Use when fine-tuning LLMs, training custom models, or adapting foundation models for specific tasks. Invoke for configuring LoRA/QLoRA adapters, preparing JSONL training datasets, setting hyperparameters for fine-tuning runs, adapter training, transfer learning, finetuning with Hugging Face PEFT, OpenAI fine-tuning, instruction tuning, RLHF, DPO, or quantizing and deploying fine-tuned models. Trigger terms include: LoRA, QLoRA, PEFT, finetuning, fine-tuning, adapter tuning, LLM training, model training, custom model.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Failed to scan

The risk profile of this skill

Quality

Content

92%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured skill body: executable examples, a sequenced workflow with validation checkpoints, and clean progressive disclosure via a reference table pointing to real bundle files. Only minor conciseness trimming is possible.

Dimension	Reasoning	Score
Conciseness	Assumes ML competence with terse, load-bearing comments and no concept-explanation padding, though the full SFTTrainer script plus a duplicated merge example is slightly heavier than a truly lean minimal example.	4 / 5
Actionability	Provides fully executable, copy-paste-ready Python covering the common cases — LoRA setup, dataset formatting, training, saving — plus a QLoRA variant and an adapter-merge snippet.	5 / 5
Workflow Clarity	Five-step Core Workflow with explicit validation checkpoints (fix all dataset errors before proceeding, monitor loss for overfitting, evaluate before deployment) covering batch training operations.	5 / 5
Progressive Disclosure	Overview body with a Reference Guide table linking to five verified one-level-deep reference files, each with a clear 'Load When' condition; bulk detail is appropriately split out.	5 / 5
	Total	19 / 20 Passed

Description

88%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description with explicit use-when guidance and a dedicated trigger-term list. It is comprehensive on capabilities and triggers, with only minor overlap risk from a few broad terms.

Dimension	Reasoning	Score
Specificity	Enumerates multiple concrete actions across the domain — configuring LoRA/QLoRA adapters, preparing JSONL datasets, setting hyperparameters, RLHF, DPO, quantizing and deploying — giving comprehensive coverage.	5 / 5
Completeness	Explicitly answers both 'what' (the enumerated actions) and 'when' via a 'Use when...' clause plus an explicit 'Trigger terms include' list.	5 / 5
Trigger Term Quality	Strong coverage of natural terms including synonyms (finetuning/fine-tuning) and core jargon (LoRA, QLoRA, PEFT, RLHF, DPO), but a few natural paraphrases (e.g. 'retrain a model', 'train on my data') are missing.	4 / 5
Distinctiveness Conflict Risk	Fine-tuning with PEFT/LoRA/RLHF/DPO is a clear specialized niche, but broad triggers like 'model training' and 'custom model' risk overlap with general ML/data skills.	4 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: jeffallan/claude-skills
Path: skills/fine-tuning-expert/SKILL.md
Commit: 882ef55

Reviewed: 4 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.