skills.rest
AlvaroBiano
AlvaroBianoCOMMUNITY
Explore all skills by this author →
GitHub

axolotl

Automate YAML-driven LLM fine-tuning with LoRA, DPO, and GRPO.

Updated May 2, 2026
Search Tags:#llm#fine-tuning#yaml#dpo#grpo#loRA#axolotl
One-click install
npx skills add https://github.com/AlvaroBiano/hermes-agent --skill axolotl-alvarobiano
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: axolotl
Source: https://github.com/AlvaroBiano/hermes-agent/tree/main/optional-skills/mlops/training/axolotl
Command: npx skills add https://github.com/AlvaroBiano/hermes-agent --skill axolotl-alvarobiano

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Axolotl enables YAML-based fine-tuning workflows for large language models, orchestrating adapters (LoRA/QLoRA), RLHF variants (DPO, ORPO, GRPO), and reward modeling to simplify setup and reproducibility.

Core Features & Use Cases

  • YAML-configured fine-tuning for multiple models and datasets with adapter management and looped evaluation.
  • Supports LoRA/QLoRA, DPO/GRPO/ORPO, KTO, and multimodal RLHF pipelines in production-grade experiments.
  • Use Case: a team wants to fine-tune a 13B LLM on a domain corpus with adapters and reward signals, then push artifacts to a model hub with minimal boilerplate.

Quick Start

Provide a YAML config describing your model, adapters, and training objectives, then run the Axolotl trainer to start fine-tuning.

Frequently Asked Questions about axolotl

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate YAML-driven fine-tuning for large language models?▼

YAML-driven fine-tuning automates LLM training by using configuration files to orchestrate adapters and RLHF variants, minimizing boilerplate for reproducible experiment pipelines.

Can I use LoRA and DPO together for multi-model RLHF workflows?▼

Yes, LoRA and DPO can be combined for multi-model RLHF workflows, alongside GRPO and ORPO, to manage adapters and integrate reward signals within a single training pipeline.

What is the best way to configure reward modeling in a production-grade LLM experiment?▼

Reward modeling in production-grade LLM experiments is configured through YAML files that define training objectives, enabling reproducible reward-model integration and adapter management.

Does YAML-configured fine-tuning work on macOS and Linux?▼

YAML-configured fine-tuning works on both macOS and Linux, enabling multi-model RLHF workflows and artifact management across these operating systems.

How do I push fine-tuned LLM adapters to a model hub after training?▼

After training, fine-tuned LLM adapters can be pushed to a model hub by defining the deployment steps in your YAML configuration, ensuring artifacts are exported with minimal boilerplate.

Related Skills

Explore More Software Engineering Skills→
devMoezdevMoez
1

axolotl

Configure Axolotl YAML for LoRA/QLoRA fine-tuning and preference training.

Community
Intermediate
linfordWulinfordWu
1

fine-tuning-with-trl

Fine-tune LLMs with TRL reinforcement learning workflows including SFT, DPO, PPO, GRPO.

Community
Advanced
ScientiaCapitalScientiaCapital
1

model-deployment

Export fine-tuned models to GGUF and deploy with vLLM or Ollama.

Community
Advanced
SignmanalSignmanal
1

axolotl

Generate validated YAML configurations for Axolotl LLM fine-tuning workflows.

Community
Intermediate