skills.rest
Rheasilvia
RheasilviaCOMMUNITY
Explore all skills by this author →
GitHub

axolotl

Fine-tune large language models via centralized YAML configuration.

Updated Jun 25, 2026
Search Tags:#llm#machine learning#fine-tuning#training#lora#axolotl
One-click install
npx skills add https://github.com/Rheasilvia/hermes-desktop --skill axolotl-rheasilvia
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: axolotl
Source: https://github.com/Rheasilvia/hermes-desktop/tree/main/optional-skills/mlops/training/axolotl
Command: npx skills add https://github.com/Rheasilvia/hermes-desktop --skill axolotl-rheasilvia

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires axolotl, torch, transformers, datasets, peft, accelerate, deepspeed, and includes references (resource) components.

What problem does it solve?

This Skill simplifies the complex process of fine-tuning Large Language Models by providing a unified, configuration-driven interface that handles everything from data preparation to advanced training techniques.

Core Features & Use Cases

  • Multi-Method Training: Supports SFT, DPO, KTO, ORPO, and GRPO training workflows.
  • Advanced Optimization: Includes built-in support for LoRA, QLoRA, FSDP, and DeepSpeed configurations.
  • Use Case: Quickly iterate on model performance by modifying a single YAML configuration file to switch between different datasets, prompt strategies, or quantization settings without rewriting training code.

Quick Start

Use the axolotl skill to initiate a supervised fine-tuning training job using the configuration file named training_config.yaml.

Frequently Asked Questions about axolotl

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

FAQPage Schema
How do I fine-tune a large language model without rewriting training scripts?▼

You can fine-tune large language models by defining parameters in a centralized YAML configuration interface. This approach handles data preparation and training pipelines without requiring code modifications.

Can I use LoRA and DeepSpeed configurations for LLM training in the same workflow?▼

Yes, LoRA and DeepSpeed configurations are supported together for LLM training. The system integrates these dependencies to manage optimization and distributed training across various hardware environments.

What training methodologies are available for fine-tuning models with this approach?▼

Available training methodologies include SFT, DPO, KTO, ORPO, and GRPO. You can switch between these workflows by modifying a single configuration file.

Does this LLM fine-tuning process support quantization settings and evaluation?▼

Yes, the fine-tuning process supports both quantization settings and evaluation pipelines. It integrates dependencies like transformers and accelerate to manage these stages.

How do I switch datasets and prompt strategies during model training?▼

You switch datasets and prompt strategies by modifying your YAML configuration file. This centralized interface enables quick iteration on model performance without rewriting training code.

What dependencies do I need to run supervised fine-tuning with torch and transformers?▼

Required dependencies include torch, transformers, datasets, peft, accelerate, and deepspeed. These libraries manage the model training, quantization, and evaluation pipelines.

Related Skills

Explore More Software Engineering Skills→
luokai25luokai25

hugging-face-tool-builder

Creates and executes scripts that interact with the Hugging Face API.

Community
Advanced
ScientiaCapitalScientiaCapital
1

model-deployment

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

Community
Advanced
longkeyylongkeyy
1

content-discovery

Automate content discovery across ArXiv, GitHub, and HuggingFace with semantic filtering and publishing.

Community
Advanced
Estrella-231Estrella-231

transformers

Run inference and fine-tuning with Hugging Face Transformers models.

Community
Intermediate