Wesley Kambale
ML Engineer | Community Lead
Mbarara, Uganda
Actions
Wesley is a machine learning engineer and data scientist, adept at crafting production-ready ML systems that provide impactful solutions in the African market.
As a tech conference speaker, he shares his expertise through insightful talks and occasional articles on TensorFlow and Keras, aiming to disseminate his knowledge and experiences.
He is also a community lead with vast experience in launching and building Google Developer communities in western Uganda. He is an active organizer in Google Developer Groups (GDG) program and an alumni of the Google Developer Students Club (GDSC) program.
Links
Area of Expertise
Topics
Keras 3: A multi-framework API for deep learning
Break free from framework limitations and explore the revolutionary world of Keras 3! This talk delves into its groundbreaking multi-framework capabilities, empowering you to leverage the strengths of TensorFlow, PyTorch, and JAX within a single, unified API. Discover how Keras 3 simplifies development, enhances flexibility, and unlocks new possibilities for your deep learning projects.
What attendees can expect:
1. Understanding the paradigm shift: Uncover the motivations and key features of Keras 3's multi-framework approach. Learn how it transcends limitations and opens doors to broader experimentation.
2. Framework flexibility in action: Witness the power of writing code that runs seamlessly across TensorFlow, PyTorch, and JAX. See how Keras 3 handles backend specifics while you focus on the core logic of your models.
3. Performance insights: Explore benchmarks and comparisons, understanding the strengths and weaknesses of each backend in different scenarios. Learn how to choose the optimal framework for your specific needs.
4. Custom component creation: Master the art of building modular layers, models, and metrics that work harmoniously across all supported frameworks. Share and reuse components effortlessly within your team.
5. Large-scale training and deployment: Discover the advanced capabilities of Keras 3 for scaling your models to handle massive datasets and deployment on diverse platforms.
Fine-tune Gemma Models using Keras and publish on Kaggle
In this workshop, we will discuss and go through a notebook on how to use Keras with Gemma open models in Google Colab and publish a fine-tuned variant of Gemma to Kaggle's model hub.
This process helps developers to build and publish their own (fine-tuned) models for quick integration and use directly from the model hub.
Understanding Diffusion Models with KerasCV
Dive into the fascinating world of Diffusion Models and unlock the power of text-to-image generation with KerasCV! This talk will unveil the inner workings of these innovative models, breaking down the complex processes into clear, understandable steps. Using KerasCV, a high-performance deep learning library, we'll explore the practical implementation of diffusion models, empowering you to create captivating images directly from your textual descriptions.
What attendees can expect:
1. Demystifying diffusion models: Gain a fundamental understanding of the core concepts behind diffusion models, including denoising, latent space, and text encoding.
2. KerasCV for effortless implementation: Discover the advantages of using KerasCV for building and training diffusion models. Witness its streamlined API and optimization techniques in action.
3. Interactive demonstrations: Witness the magic firsthand! We'll walk through live coding examples, generating breathtaking images based on text prompts right before your eyes.
4. Beyond the basics: Explore advanced applications of diffusion models, including image editing, inpainting, and style transfer.
5. Actionable insights: Uncover practical tips and tricks to enhance your own image generation experiments with KerasCV.
Keras for the Young and Curious
For the eager beginners looking to start a career in machine learning and deep learning and for the people who want to start building right away.
The session covers the use cases of Keras such as KerasNLP - for Natural Language Processing, and KerasCV - for Computer Vision. It also covers insights on models such as StableDiffusion (v1 & 2), YOLO, CSPDarknet, EfficientNet, etc.
The session is introductory in nature and is meant for beginners and intermediate.
LLM on Android with Keras and TensorFlow Lite
This session involves practical and theoretical demonstrations of how to train and deploy your own large language model (LLM) on Android.
The session tackles ML problem framing such as text generation using a fine-tuned pre-trained GPT-2 model.
We explore the use of datasets using Dataset Search to prepare data for our LLM. We create a TensorFlow dataset and run preprocessing on it. The tf.data API helps to map the gpt2_preprocessor against the training dataset.
We train an LLM using pre-trained models and deploy them to Android using TensorFlow Serving.
This session assumes that the learner is aware of mobile technologies like Android and Flutter and also aware of Natural Language Processing.
Understanding Neural Networks with Keras and TensorFlow
The session focuses on the a deep and practical understanding of deep learning's neural networks using Keras and TensorFlow for developers to build better and efficient ML models.
The session takes into account the mathematics and building blocks behind neural networks and demonstrates to the attendees how to build one and make it efficient through the concepts of fine-tuning and transfer learning.
KerasCV for the Young and Curious
For the eager beginners looking to start with Computer Vision and for the people who want to start building right away.
The session covers the import use cases and models such as StableDiffusion (v1 & 2), YOLO, CSPDarknet, EfficientNet, etc. that can be used in CV.
The session is introductory in nature and is meant for beginners and intermediate.
Automation & Integration Using Python
This session discusses the libraries and tools that make it easy to automate repetitive tasks, integrate different systems and applications, and streamline workflows in Python.
We explore various aspects of automation and integration using Python, with examples and common use cases such as web scraping, API integration, database integration, creating automated reports, etc.
Session slides: https://speakerdeck.com/kambale/automation-and-integration-using-python
PyCon Uganda 2024 Upcoming
- Serving TensorFlow Models with the Fast API
I/O Extended Mbarara 2024 Sessionize Event
IWD Jinja 2024 #ImpactTheFuture Sessionize Event
International Women's Day Mbarara 2024 Sessionize Event
Global AI Conference 2023 Sessionize Event
DevFest Kampala 2023 Sessionize Event
DevFest Mbarara 2023 Sessionize Event
PyCon Uganda 2023 Sessionize Event
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top