Training programme overview
Advanced specialization on open‑source models: model selection and adaptation, dataset preparation, PEFT/LoRA/QLoRA/TRL methods, fine‑tuning and evaluation for deployment.
Maps the open‑source model ecosystem: model families, properties (size, memory, latency) and selection by use case.
Covers dataset creation and management: cleaning, labeling, splits, evaluation metrics and experiment pipelines.
Describes fine‑tuning and adaptation techniques: PEFT, LoRA/QLoRA, tuning strategies and tools for efficient training of open models.
Addresses packaging and deployment: conversions, inference optimizations, post‑deploy monitoring and continuous evaluation of production models.
Training · Advanced
Courses included in this training programme
Open-Source Model Ecosystem
Survey of the open-source model ecosystem: model families, licensing and use cases; model hubs and model cards; Transformers pipelines and tokenizers; and how to choose between API, local hosting or fine-tuning for LLM deployment.
4 lessons · 24 activities · 15 exercises · 2 videos
Datasets and Evaluation
Design and evaluation of datasets for fine-tuning: instruction-tuning formatting, train/validation/test split strategies, baseline evaluation before fine-tuning, and checks for bias, leakage and data quality.
4 lessons · 20 activities · 0 exercises · 3 videos
Fine-Tuning Techniques
Comparison of fine‑tuning approaches for open‑source LLMs, detailing full fine‑tuning versus PEFT, deep dives into LoRA/QLoRA and the TRL SFTTrainer, and an examination of hyperparameters and common failure modes to monitor during tuning.
4 lessons · 24 activities · 13 exercises · 3 videos
Packaging and Deployment
Packaging and production readiness for fine‑tuned models: saving and merging adapters, writing model cards and dataset documentation, serving strategies, and cost, governance and lifecycle management considerations.
4 lessons · 24 activities · 16 exercises · 2 videos
Access the platform
Open the catalogue to review course details, access conditions, and your personal progress. Sign in.