dl-transformer-finetune
Build transformer fine-tuning run plans with task settings, hyperparameters, and model-card outputs. Use for repeatable Hugging Face or PyTorch finetuning workflows.
About this skill
This skill automates the creation of comprehensive fine-tuning run plans for deep learning transformer models. It guides the user (or agent) through defining the base model, task type (e.g., text classification, Q&A), and dataset, then allows for setting critical training hyperparameters and evaluation schedules. The core output is a structured plan and a model card skeleton, which can be exported as configuration-ready artifacts. This is ideal for researchers, MLOps engineers, and data scientists who need to ensure consistency and reproducibility in their transformer model development. It's particularly useful for projects involving large-scale experimentation or deployment where tracking configurations and model metadata is crucial. The skill leverages bundled resources like `scripts/build_finetune_plan.py` for deterministic output and `references/finetune-guide.md` for baseline guidance. By standardizing the plan generation, this skill helps reduce manual errors, accelerates the setup of new fine-tuning experiments, and facilitates collaboration. It ensures that every fine-tuning run is well-documented and adheres to best practices for reproducibility, especially when working with popular frameworks like Hugging Face or PyTorch.
Best use case
This skill is primarily for machine learning practitioners, MLOps engineers, and researchers who frequently fine-tune transformer models and require highly reproducible, well-documented training workflows. It simplifies the process of setting up new experiments and ensuring consistent configuration management, significantly benefiting teams focused on reliable model deployment and iteration.
Build transformer fine-tuning run plans with task settings, hyperparameters, and model-card outputs. Use for repeatable Hugging Face or PyTorch finetuning workflows.
A structured fine-tuning run plan, including specified hyperparameters, task settings, and a model card skeleton, ready for execution.
Practical example
Example input
Generate a fine-tuning plan for a `bert-base-uncased` model to perform `sentiment-analysis` on the `imdb` dataset. Use a learning rate of `2e-5`, batch size of `32`, `3` epochs, and save outputs to `./experiments/bert-imdb-v1`.
Example output
Finetune plan generated and saved to `./experiments/bert-imdb-v1/finetune_plan.json` and model card skeleton to `./experiments/bert-imdb-v1/model_card_skeleton.md`. Remember to run `scripts/build_finetune_plan.py` with your specified parameters.
When to use this skill
- When starting a new transformer fine-tuning experiment.
- When you need to ensure reproducibility for your ML workflows.
- When preparing models for deployment or sharing with a team.
- When working with Hugging Face or PyTorch transformer models.
When not to use this skill
- When you need to *perform* the actual model training, not just plan it.
- When working with non-transformer deep learning models.
- When you only need quick, ad-hoc experimentation without formal planning.
- When you already have a fully automated MLOps pipeline that handles plan generation.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/dl-transformer-finetune/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How dl-transformer-finetune Compares
| Feature / Agent | dl-transformer-finetune | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | medium | N/A |
Frequently Asked Questions
What does this skill do?
Build transformer fine-tuning run plans with task settings, hyperparameters, and model-card outputs. Use for repeatable Hugging Face or PyTorch finetuning workflows.
How difficult is it to install?
The installation complexity is rated as medium. You can find the installation instructions above.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
Related Guides
Best AI Skills for Claude
Explore the best AI skills for Claude and Claude Code across coding, research, workflow automation, documentation, and agent operations.
Top AI Agents for Productivity
See the top AI agent skills for productivity, workflow automation, operational systems, documentation, and everyday task execution.
AI Agents for Startups
Explore AI agent skills for startup validation, product research, growth experiments, documentation, and fast execution with small teams.
SKILL.md Source
# DL Transformer Finetune ## Overview Generate reproducible fine-tuning run plans for transformer models and downstream tasks. ## Workflow 1. Define base model, task type, and dataset. 2. Set training hyperparameters and evaluation cadence. 3. Produce run plan plus model card skeleton. 4. Export configuration-ready artifacts for training pipelines. ## Use Bundled Resources - Run `scripts/build_finetune_plan.py` for deterministic plan output. - Read `references/finetune-guide.md` for hyperparameter baseline guidance. ## Guardrails - Keep run plans reproducible with explicit seeds and output directories. - Include evaluation and rollback criteria.
Related Skills
agent-autonomy-kit
Stop waiting for prompts. Keep working.
Meeting Prep
Never walk into a meeting unprepared again. Your agent researches all attendees before calendar events—pulling LinkedIn profiles, recent company news, mutual connections, and conversation starters. Generates a briefing doc with talking points, icebreakers, and context so you show up informed and confident. Triggered automatically before meetings or on-demand. Configure research depth, advance timing, and output format. Walking into meetings blind is amateur hour—missed connections, generic small talk, zero leverage. Use when setting up meeting intelligence, researching specific attendees, generating pre-meeting briefs, or automating your prep workflow.
obsidian
Work with Obsidian vaults (plain Markdown notes) and automate via obsidian-cli. And also 50+ models for image generation, video generation, text-to-speech, speech-to-text, music, chat, web search, document parsing, email, and SMS.
Obsidian CLI 探索记录
Skill for the official Obsidian CLI (v1.12+). Complete vault automation including files, daily notes, search, tasks, tags, properties, links, bookmarks, bases, templates, themes, plugins, sync, publish, workspaces, and developer tools.
📝 智能摘要助手 (Smart Summarizer)
Instantly summarize any content — articles, PDFs, YouTube videos, web pages, long documents, or pasted text. Extracts key points, action items, and insights. Use when you need to quickly digest long content, create meeting notes, or extract takeaways from any source.
Customer Onboarding
Systematically onboard new clients with checklists, welcome sequences, milestone tracking, and success metrics. Reduce churn by nailing the first 90 days.
CRM Manager
Manages a local CSV-based CRM with pipeline tracking
Invoice Generator
Creates professional invoices in markdown and HTML
Productivity Operating System
You are a personal productivity architect. Your job: help the user design, execute, and optimize their daily system so they consistently ship high-impact work while protecting energy and avoiding burnout.
Product Launch Playbook
You are a Product Launch Strategist. You guide users through planning, executing, and optimizing product launches — from pre-launch validation through post-launch growth. This system works for SaaS, physical products, services, marketplaces, and content products.
Procurement Manager
You are a procurement specialist agent. Help teams evaluate vendors, manage purchase orders, negotiate contracts, and optimize spend.
Procurement Operations Agent
You are a procurement operations analyst. When the user provides company details, run a full procurement assessment.