Build Large Language Models (LLM) and Deep Learning Applications with Keras 3 Training Course and Workshop in Bangalore, Mysore, Chennai, Hyderabad, Pune, Mumbai, Delhi, Noida, Gurgaon, Kolkata

Program Overview
Pre-Requisites

This Build Large Language Models (LLM) and Deep Learning Applications with Keras 3 training course is an all-encompassing program for AI practitioners aiming to advance their skills in natural language processing and computer vision. Starting with an overview of Keras 3’s versatile architecture, participants will dive into backend integrations with TensorFlow, PyTorch, and JAX, explore KerasHub for pretrained models, and leverage the latest features in KerasNLP and KerasCV. This robust foundation enables the development of sophisticated language and vision applications.

This course covers essential deep learning techniques, from creating transformer-based LLMs and custom chatbots to fine-tuning models for real-world applications, including hyperparameter tuning and model parallelism. By the end, learners will be able to deploy multimodal applications that integrate both language and vision, effectively preparing them for high-impact roles in AI and machine learning.

Attendees for Build Large Language Models (LLM) and Deep Learning Applications with Keras 3 workshop should have:

Minimum of 1 year of working experience.

Introduction to Keras, Large Language Models, and Backend Integration

Overview of Keras 3 features and high-level architecture
Integrating Keras backends - TensorFlow, PyTorch, JAX
Introduction to KerasHub for pretrained resources
Overview of KerasCV and KerasNLP libraries

Fundamentals of Deep Learning with Keras

Basic components: layers, activations, model definitions
Building foundational LLMs: recurrent and transformer models
Transformer architectures and Keras’ multi-backend API
Loading and using pretrained models in KerasNLP (e.g., Gemma)

Chatbot Design and Conversational AI

Using KerasNLP for basic chatbot design
Techniques for maintaining conversational context with prompts
Implementing conversational markup for turn-by-turn chat
Building interactive applications and customizing responses

Training and Fine-Tuning Large Language Models

Fine-tuning techniques for LLMs (e.g., chatbot to formal or informal tone)
Techniques like LoRA (Low-Rank Adaptation) for efficient fine-tuning
Enabling model parallelism with Keras to handle large models
Hyperparameter tuning and resource management with KerasTuner

NLP Applications Using KerasNLP

Loading instruction-tuned models for NLP applications (summarization, Q&A, text generation)
Task-specific fine-tuning using domain-specific data (e.g., car sales language)
Managing multi-turn conversations and context length handling
Deploying NLP models for real-world applications with KerasHub

Advanced Fine-Tuning Techniques and Model Customization

Introduction to control vectors for changing model behavior without retraining
Using PCA-based techniques to adjust tone and model response
Extending model context length with the self-extend technique for longer prompts
Detailed implementation of control vectors and custom response adjustments

Computer Vision Applications Using KerasCV

Basics of image classification, detection, and segmentation in KerasCV
Developing multimodal (vision-NLP) applications with Keras
Preprocessing, augmentation, and KerasHub resources for vision models
Fine-tuning and deployment of KerasCV models

Encarta Labs Advantage

One Stop Corporate Training Solution Providers for over 6,000 various courses on a variety of subjects
All courses are delivered by Industry Veterans
Get jumpstarted from newbie to production ready in a matter of few days

Trained more than 50,000 Corporate executives across the Globe
All our trainings are conducted in workshop mode with more focus on hands-on sessions

Build Large Language Models (LLM) and Deep Learning Applications with Keras 3

COURSE AGENDA