Fine-Tuning Large Language Models: From Custom Datasets to High-Performance AI Models Using Modern Toolchains - Softcover

Inhaltsangabe

Are your prompts not cutting it anymore? Fine-tuning is the next level.

This hands-on guide takes you from zero to deploying your own
custom-trained large language model using real code, modern tools,
and production-ready techniques used by AI engineers today.

Whether you want to build a medical assistant, legal document analyzer,
coding AI, or customer support bot, this book gives you the exact
pipeline to make it happen.

What you'll learn:

- How fine-tuning differs from prompt engineering and RAG and when
to use each
- LoRA and QLoRA: fine-tune powerful models on a single GPU
- How to collect, clean, and format high-quality instruction datasets
- Supervised Fine-Tuning (SFT), RLHF, and Direct Preference Optimization
- Advanced techniques: mixed precision, gradient checkpointing,
multi-GPU training
- How to evaluate, quantize, and deploy your model to production
- Safety, bias mitigation, and responsible AI practices
- 14 chapters of step-by-step projects including a chatbot,
evaluation dashboard, and automated training pipeline

This is not a theory book. Every chapter includes working Python code
using Hugging Face Transformers, PEFT, TRL, and modern open-source
models like Llama and Mistral.

By the end, you won't just understand fine-tuning you'll have built
and shipped a model of your own.

Perfect for: ML engineers, Python developers, data scientists, and
anyone ready to move beyond ChatGPT and take control of their AI stack.

Die Inhaltsangabe kann sich auf eine andere Ausgabe dieses Titels beziehen.