Aulafy
ESEN
Courses
Home/Courses/Fine-tuning and post-training for LLMs
capsule/fine-tuning-local

Fine-tuning and post-training for LLMs

Prepare datasets, run LoRA and QLoRA jobs, evaluate overfitting and export models for local use.

Intermediate to advanced 8 lessons Free and open source
Start lessons

Lessons are available in English. Code snippets keep original commands and file names intact.

modules
3
lessons
8
status
open
Roadmap

Syllabus

8 steps

module 01 / Before training

2 lessons
  1. 1Map: SFT, LoRA, QLoRA, and DPO/fine-tuning-local/mapa-post-training
  2. 2High-quality instruction datasets/fine-tuning-local/datasets-instrucciones

module 02 / Practical training

3 lessons
  1. 3LoRA and QLoRA without the hype/fine-tuning-local/lora-qlora
  2. 4Fast SFT with Unsloth/fine-tuning-local/unsloth-sft
  3. 5Axolotl for reproducible training/fine-tuning-local/axolotl-config

module 03 / Evaluate and use

3 lessons
  1. 6Evals, Overfitting, and Regressions/fine-tuning-local/evals-overfitting
  2. 7Export to GGUF and Ollama/fine-tuning-local/export-gguf-ollama
  3. 8Project: adapted model for an SMB/fine-tuning-local/proyecto-modelo-pyme
View all courses