Tool or Technology
TT0176

Mastering Hugging Face

Advanced Model Engineering, Fine-Tuning, Agents & Enterprise Deployment

Hugging Face
Duration
32 Hours
Level
Advanced
Delivery
Instructor-Led
Format
Capability Development

Programme Objectives

  • Develop advanced expertise in discovering, evaluating, using, adapting, documenting, and managing AI models and datasets through the Hugging Face ecosystem.
  • Build production-ready AI solutions using Transformers, Datasets, Diffusers, Sentence Transformers, Accelerate, PEFT, TRL, and evaluation frameworks.
  • Fine-tune and post-train foundation models using supervised fine-tuning, LoRA/PEFT, preference optimization, reward-based techniques, and distributed-compute strategies.
  • Deploy scalable AI applications using Inference Providers, dedicated Inference Endpoints, Hugging Face Jobs, Spaces, and programmatic Hub services.
  • Build tool-using and multi-step AI Agents using smolagents, RAG, external tools, MCP, secure execution, and human-in-the-loop patterns.
  • Establish enterprise controls for model licensing, dataset governance, access permissions, SSO, tokens, gated repositories, resource groups, audit logs, network security, and responsible AI.

Technology covered

Hugging Face HubTransformersDatasetsTokenizersSentence TransformersDiffusersAcceleratePEFTLoRATRLEvaluateSafetensorshuggingface_hubHF CLIInferenceClientInference ProvidersInference EndpointsHugging Face JobsSandboxesSpacesGradiosmolagentsMCPModel CardsDataset CardsOrganizationsResource Groups & Enterprise Hub

Who should attend

  • AI Engineers
  • Machine Learning Engineers
  • Generative AI Engineers
  • Data Scientists
  • NLP Engineers
  • Computer Vision Engineers
  • MLOps Engineers
  • AI Application Developers
  • Software Developers
  • Research Engineers
  • Solution Architects
  • AI / ML Platform Engineers
  • Model Fine-Tuning Teams
  • AI Centre of Excellence Teams
  • Enterprise AI & Innovation Teams

Prerequisites & Participant Readiness

  • Working knowledge of Python
  • Basic understanding of machine learning and deep learning
  • Familiarity with neural networks and transformer concepts is recommended
  • Basic knowledge of PyTorch is beneficial
  • Familiarity with APIs, JSON, Git, and command-line tools is recommended
  • Basic understanding of Generative AI and Large Language Models is helpful
  • Access to suitable CPU/GPU compute is recommended for fine-tuning exercises
  • Basic cloud and container knowledge is beneficial for deployment modules
  • Some Enterprise Hub, dedicated compute, Jobs, Spaces hardware, and security features depend on subscription and organisation configuration
  • smolagents and Hugging Face Sandboxes include experimental capabilities that may evolve
  • Completion of shorter Hugging Face programmes is not required

TOC Modules

Concepts
  • Understanding Hugging Face as an ecosystem for models, datasets, applications, training, inference, Agents, and collaboration
  • Understanding the relationship between the Hub and major Hugging Face libraries
  • Understanding pretrained foundation models versus task-specific models
  • Understanding open, gated, private, and commercially licensed model access
  • Understanding end-to-end AI lifecycle architecture
Practical activities
  • Exploring models, datasets, Spaces, and organisations on the Hub
  • Comparing multiple models for the same business requirement
  • Analysing model metadata, licence, architecture, task, and deployment options
  • Mapping Business Requirement → Model → Data → Adaptation → Evaluation → Deployment

Scenarios

Enterprise Domain-Specific AI Assistant

Approved Enterprise Data → Hugging Face Datasets → Foundation Model Selection → Transformers → PEFT / LoRA → Evaluation → Model Card → Inference Endpoint → RAG / smolagents → Business Application → Monitoring

Participants build a domain-specific AI assistant by preparing governed data, adapting an appropriate foundation model, evaluating it against business criteria, deploying it through managed inference, and connecting it to organisational knowledge and Agent tools.

Governed Open-Model AI Platform

Business Requirement → Hub Model Evaluation → Licence & Security Review → Enterprise Repository → Hugging Face Jobs → Fine-Tuning / TRL → Evaluation → Inference Endpoint → Space / API → Resource Groups → Audit & Governance

Participants architect an enterprise model platform that enables teams to discover, adapt, test, deploy, and reuse AI models while maintaining central controls for access, security, licences, model documentation, compute, and production deployment.

Related programmes

Continue with programmes from the same capability area.

Take the next step

Ready to make this programme work for your team?

Customise modules, duration and business scenarios for your team.

Instructor-ledVirtualHybrid

Designed around your roles, tools and real workflows.