#
ENJA

Jackrong-llm-finetuning-guide

R6410418/Jackrong-llm-finetuning-guide

An educational knowledge base and collection of recipes for LLM fine-tuning, reinforcement learning, and model conversion.

1.6k 263Apache-2.0Updated 2026-07-11

Overview

This repository provides comprehensive guidance, code scripts, and notebooks for fine-tuning open-source large language models using techniques like LoRA, QLoRA, SFT, GRPO, and GSPO. It includes training recipes for Google Colab and Kaggle, dataset distillation tools, curated datasets, and specific workflows for Qwen MTP GGUF conversion and quantization.

Capabilities

  • SFT with LoRA/QLoRA training scripts
  • GRPO and GSPO reinforcement learning tutorials
  • Dataset distillation and batch download helpers
  • Qwen MTP GGUF conversion and validation pipeline

Best for

Fine-tuning large language models on custom datasets, Executing reinforcement learning workflows (GRPO and GSPO), Preparing, filtering, and distilling training data, Converting and quantizing models to GGUF format with MTP tensor support

Works with

CodexClaude Code
Evoa Score breakdown= Σ (score × weight)
72
Task usefulness20%77 → +15.4

実タスクにどれだけ役立つか(機能の豊富さ・用途の明確さ)。 AIによるcapabilities/use-cases解析

Code quality15%85 → +12.8

実装・指示の品質。 AIによるSKILL.md/README解析

Maintenance15%85 → +12.8

リポジトリがどれだけ活発に保守されているか。 GitHub 最終push日時の新しさ

Documentation12%24 → +2.9

ドキュメントの充実度・分かりやすさ。 README/独自要約の情報量

Security15%80 → +12.0

危険・不審な挙動が無いか。 AIによるセキュリティレビュー

Originality10%75 → +7.5

ありふれたラッパーではない独自性。 AIによる独自性判定

Popularity8%83 → +6.6

コミュニティの採用度。 GitHub Stars/Forks(対数スケール)

Compatibility5%50 → +2.5

対応AIエージェントの広さ。 AIによる対応エージェント判定

Weighted total72.4 / 100

ライセンス不明/制限あり(Red)のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →

Security considerations

Ensure no private API keys, sensitive tokens, or proprietary datasets are committed to training configurations or logs.

Categories

Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.