Advanced Deep Learning Architectures and Transformers Course

Advance your expertise in transformers, multimodal architectures, efficient fine-tuning, scaling, ablation and production inference optimisation.

Artificial Intelligence and Machine Learning Courses Confirmed

Training Locations

This Advanced Deep Learning Architectures and Transformers Course is available in multiple cities. Please select your preferred location from the list below

Durrës

Albania

Tirana

Albania

Andorra la Vella

Andorra

Escaldes-Engordany

Andorra

Innsbruck

Austria

Salzburg

Austria

Vienna

Austria

Gomel

Belarus

Minsk

Belarus

Antwerp

Belgium

Brussels

Belgium

Banja Luka

Bosnia and Herzegovina

Sarajevo

Bosnia and Herzegovina

Plovdiv

Bulgaria

Sofia

Bulgaria

Dubrovnik

Croatia

Split

Croatia

Zagreb

Croatia

Limassol

Cyprus

Nicosia

Cyprus

Brno

Czech Republic

Prague

Czech Republic

Aarhus

Denmark

Copenhagen

Denmark

Tallinn

Estonia

Tartu

Estonia

Helsinki

Finland

Tampere

Finland

Lyon

France

Marseille

France

Nice

France

Paris

France

Berlin

Germany

Frankfurt

Germany

Hamburg

Germany

Munich

Germany

Athens

Greece

Thessaloniki

Greece

Budapest

Hungary

Debrecen

Hungary

Akureyri

Iceland

Reykjavík

Iceland

Cork

Ireland

Dublin

Ireland

Florence

Italy

Milan

Italy

Naples

Italy

Rome

Italy

Pristina

Kosovo

Prizren

Kosovo

Liepāja

Latvia

Riga

Latvia

Schaan

Liechtenstein

Vaduz

Liechtenstein

Kaunas

Lithuania

Vilnius

Lithuania

Esch-sur-Alzette

Luxembourg

Luxembourg City

Luxembourg

St. Julian's

Malta

Valletta

Malta

Bălți

Moldova

Chișinău

Moldova

La Condamine

Monaco

Monte Carlo

Monaco

Budva

Montenegro

Podgorica

Montenegro

Amsterdam

Netherlands

Rotterdam

Netherlands

The Hague

Netherlands

Ohrid

North Macedonia

Skopje

North Macedonia

Bergen

Norway

Oslo

Norway

Gdańsk

Poland

Kraków

Poland

Warsaw

Poland

Faro

Portugal

Lisbon

Portugal

Porto

Portugal

Bucharest

Romania

Cluj-Napoca

Romania

City of San Marino

San Marino

Serravalle

San Marino

Belgrade

Serbia

Novi Sad

Serbia

Bratislava

Slovakia

Košice

Slovakia

Bled

Slovenia

Ljubljana

Slovenia

Barcelona

Spain

Madrid

Spain

Valencia

Spain

Gothenburg

Sweden

Stockholm

Sweden

Bern

Switzerland

Geneva

Switzerland

Zurich

Switzerland

Kyiv

Ukraine

Lviv

Ukraine

Odesa

Ukraine

Dubai

United Arab Emirates

Birmingham

United Kingdom

Edinburgh

United Kingdom

London

United Kingdom

Manchester

United Kingdom

Rome (Vatican-adjacent)

Vatican City

Training Outlines

Introduction

Modern deep-learning systems depend on architectural and optimisation choices that extend well beyond standard feed-forward networks. Attention mechanisms, transformers, multimodal models and large-scale pretraining introduce challenges involving gradient flow, memory complexity, adaptation, distributed training and inference efficiency. Advanced practitioners must be able to analyse these trade-offs rather than rely on framework defaults.

This five-day advanced course is designed for professionals already familiar with neural-network training. Participants will examine transformer internals, modern architecture patterns, efficient adaptation, large-scale training and production optimisation. Technical laboratories emphasise controlled experimentation, ablation and evidence-based architecture selection for demanding language, vision and multimodal applications.

Objectives

By the end of this course, participants will be able to:

Diagnose optimisation and gradient-flow problems in deep networks.
Analyse transformer components and computational complexity.
Design encoder, decoder and multimodal architectures.
Apply efficient fine-tuning and adaptation methods.
Optimise memory, throughput and inference latency.
Evaluate advanced models through ablation and robustness testing.
Select architectures against technical and operational constraints.
Defend an advanced deep-learning system design.

Course Content

Day 1: Advanced Architecture and Optimisation

Residual, dense and gated connections
Gradient flow and initialisation
Normalisation architecture and placement
Adaptive optimisation and learning schedules
Regularisation and stochastic depth
Deep-network diagnostic laboratory

Day 2: Transformer Internals

Scaled dot-product and multi-head attention
Positional, relative and rotary encoding
Pre-norm and post-norm transformer blocks
Causal masking and autoregressive decoding
KV caching and attention complexity
Transformer-component implementation laboratory

Day 3: Modern Transformer Architectures

Encoder, decoder and sequence-to-sequence designs
Vision transformers and patch representations
Cross-attention and multimodal fusion
Sparse attention and long-context methods
Mixture-of-experts architecture
Architecture-selection design review

Day 4: Scaling, Adaptation and Evaluation

Self-supervised and pretraining objectives
Mixed-precision and distributed training
Adapters, LoRA and parameter-efficient tuning
Quantisation, pruning and distillation
Ablation, robustness and calibration
Efficient adaptation laboratory

Day 5: Advanced Inference and Deployment

Dynamic batching and inference optimisation
Model routing and serving architecture
Retrieval versus parameter adaptation
Interpretability and failure analysis
Monitoring, safety and change control
Capstone transformer-system defence

Training Schedule

Below is the table of cities along with the respective dates for the upcoming training sessions of Advanced Deep Learning Architectures and Transformers Course. Please review the schedule to find the most convenient option for you. You can also use the below search bar to type the city name and filter the results.