Formation en IA & Data: LLM Fine-Tuning – Adapting a Model to Your Business Data – Advanced - Ascent Formation
Back to trainings
IA & Data

LLM Fine-Tuning – Adapting a Model to Your Business Data – Advanced

3 jour(s)21h

Description

Master advanced techniques for fine-tuning open-source large language models to adapt an LLM to a specific business domain. The training covers the construction of a complete fine-tuning pipeline, including model selection, data preparation, training with PEFT methods (LoRA, QLoRA), performance evaluation, and deployment of a specialized model.

Learning Objectives

  • Understand the architectures and limitations of major open-source LLMs.
  • Select an adaptation strategy suited to a business use case.
  • Prepare a training dataset for a specialized LLM.
  • Implement fine-tuning using Hugging Face Transformers and PEFT.
  • Use LoRA and QLoRA techniques to optimize training.
  • Evaluate the performance of a fine-tuned model.
  • Deploy a specialized model in an application environment.

Target Audience

Data scientists
Machine learning engineers
NLP engineers
AI engineers
Data/IA architects
Developers specialized in generative AI

Prerequisites

Strong proficiency in Python
Knowledge of machine learning and deep learning
Experience with Transformer architectures
Familiarity with the Hugging Face ecosystem

Program Outline

Informations

Duration

3 jour(s)

21h

Tarif

Sur demande

    LLM Fine-Tuning – Adapting a Model to Your Business Data – Advanced | Ascent Formation | Ascent Formation