Twin Cities Deep Learning Study Group

Intro to Deep Learning (Using Deep Learning with Python, 3rd Ed.)

Curious about deep learning but don’t know where to start? Join our in-person, hands-on meetup at the library as we work through François Chollet’s brand-new Deep Learning with Python, Third Edition. We’ll demystify neural nets from first principles and build real models together using Python. The book is a complete, modern rewrite that now covers Keras 3, TensorFlow, PyTorch, JAX, and today’s generative AI techniques—all taught by the creator of Keras. 

What we’ll cover (highlights from the new 3rd edition)
• Foundations: core deep learning ideas and a code-first, intuitive approach. 
• Modern tooling: Keras 3 across multiple backends, with primers on TensorFlow, PyTorch, and JAX. 
• Computer vision: image classification, image segmentation, and an intro to object detection. 
• Timeseries: practical forecasting workflows. 
• NLP & LLMs: text classification, machine translation, and large language models—including building a small GPT-like model and understanding sampling/decoding. 
• Generative images: diffusion-based image generation (the tech behind modern AI art tools). 
• Scaling & tuning: tips for faster training, multi-GPU/TPU, and model optimization. 
• Bigger & better than before: the third edition adds ~30% more content and reflects the 2024–2025 state of the art. 

Who it’s for
• Python-comfortable beginners to ML (you don’t need prior machine learning or heavy math). 

How we’ll learn
• Short explanations + live coding in Python each session.
• We’ll follow the book’s examples (with cross-references to PyTorch/JAX where helpful).
• Optional: run the official companion notebooks from the authors’ GitHub. 

Getting the book
• You can read the 3rd edition online for free on the official site, with runnable code links in each chapter. https://deeplearningwithpython.io/
• Release details: Manning lists the print/ePub/liveBook availability as September 30, 2025; Amazon lists the Kindle edition publication date as October 14, 2025. 

What to bring
• Laptop with Python 3 and a Jupyter environment. We’ll start with Keras 3 + TensorFlow; PyTorch/JAX are welcome too. (We’ll share an environment setup link at the start.) 

Location & vibe
• Where: [Library Meeting Room], with tables, power, and Wi-Fi.
• Format: 15-minute mini-lecture → 60-minute guided coding → 15-minute Q&A/show-and-tell.
• Experience level: friendly to newcomers; helpful for practitioners who want a 2025 refresh.

First session agenda (75–90 minutes)

1. Why deep learning now? The Keras 3 ecosystem in 2025. 
2. Your first neural network in Keras (classification).
3. What’s next: where vision, timeseries, LLMs, and diffusion fit in our roadmap. 

RSVP to save your seat—bring your curiosity and a laptop. Let’s learn by building!

Discord server: https://discord.gg/WVRpVGUH

Intro to Deep Learning (Using Deep Learning with Python, 3rd Ed.)

Curious about deep learning but don’t know where to start? Join our in-person, hands-on meetup at the library as we work through François Chollet’s brand-new Deep Learning with Python, Third Edition. We’ll demystify neural nets from first principles and build real models together using Python. The book is a complete, modern rewrite that now covers Keras 3, TensorFlow, PyTorch, JAX, and today’s generative AI techniques—all taught by the creator of Keras. 

What we’ll cover (highlights from the new 3rd edition)
• Foundations: core deep learning ideas and a code-first, intuitive approach. 
• Modern tooling: Keras 3 across multiple backends, with primers on TensorFlow, PyTorch, and JAX. 
• Computer vision: image classification, image segmentation, and an intro to object detection. 
• Timeseries: practical forecasting workflows. 
• NLP & LLMs: text classification, machine translation, and large language models—including building a small GPT-like model and understanding sampling/decoding. 
• Generative images: diffusion-based image generation (the tech behind modern AI art tools). 
• Scaling & tuning: tips for faster training, multi-GPU/TPU, and model optimization. 
• Bigger & better than before: the third edition adds ~30% more content and reflects the 2024–2025 state of the art. 

Who it’s for
• Python-comfortable beginners to ML (you don’t need prior machine learning or heavy math). 

How we’ll learn
• Short explanations + live coding in Python each session.
• We’ll follow the book’s examples (with cross-references to PyTorch/JAX where helpful).
• Optional: run the official companion notebooks from the authors’ GitHub. 

Getting the book
• You can read the 3rd edition online for free on the official site, with runnable code links in each chapter. https://deeplearningwithpython.io/
• Release details: Manning lists the print/ePub/liveBook availability as September 30, 2025; Amazon lists the Kindle edition publication date as October 14, 2025. 

What to bring
• Laptop with Python 3 and a Jupyter environment. We’ll start with Keras 3 + TensorFlow; PyTorch/JAX are welcome too. (We’ll share an environment setup link at the start.) 

Location & vibe
• Where: [Library Meeting Room], with tables, power, and Wi-Fi.
• Format: 15-minute mini-lecture → 60-minute guided coding → 15-minute Q&A/show-and-tell.
• Experience level: friendly to newcomers; helpful for practitioners who want a 2025 refresh.

First session agenda (75–90 minutes)

1. Why deep learning now? The Keras 3 ecosystem in 2025. 
2. Your first neural network in Keras (classification).
3. What’s next: where vision, timeseries, LLMs, and diffusion fit in our roadmap. 

RSVP to save your seat—bring your curiosity and a laptop. Let’s learn by building!

Discord server: https://discord.gg/WVRpVGUH

Intro to Deep Learning (Using Deep Learning with Python, 3rd Ed.)

Curious about deep learning but don’t know where to start? Join our in-person, hands-on meetup at the library as we work through François Chollet’s brand-new Deep Learning with Python, Third Edition. We’ll demystify neural nets from first principles and build real models together using Python. The book is a complete, modern rewrite that now covers Keras 3, TensorFlow, PyTorch, JAX, and today’s generative AI techniques—all taught by the creator of Keras. 

What we’ll cover (highlights from the new 3rd edition)
• Foundations: core deep learning ideas and a code-first, intuitive approach. 
• Modern tooling: Keras 3 across multiple backends, with primers on TensorFlow, PyTorch, and JAX. 
• Computer vision: image classification, image segmentation, and an intro to object detection. 
• Timeseries: practical forecasting workflows. 
• NLP & LLMs: text classification, machine translation, and large language models—including building a small GPT-like model and understanding sampling/decoding. 
• Generative images: diffusion-based image generation (the tech behind modern AI art tools). 
• Scaling & tuning: tips for faster training, multi-GPU/TPU, and model optimization. 
• Bigger & better than before: the third edition adds ~30% more content and reflects the 2024–2025 state of the art. 

Who it’s for
• Python-comfortable beginners to ML (you don’t need prior machine learning or heavy math). 

How we’ll learn
• Short explanations + live coding in Python each session.
• We’ll follow the book’s examples (with cross-references to PyTorch/JAX where helpful).
• Optional: run the official companion notebooks from the authors’ GitHub. 

Getting the book
• You can read the 3rd edition online for free on the official site, with runnable code links in each chapter. https://deeplearningwithpython.io/
• Release details: Manning lists the print/ePub/liveBook availability as September 30, 2025; Amazon lists the Kindle edition publication date as October 14, 2025. 

What to bring
• Laptop with Python 3 and a Jupyter environment. We’ll start with Keras 3 + TensorFlow; PyTorch/JAX are welcome too. (We’ll share an environment setup link at the start.) 

Location & vibe
• Where: [Library Meeting Room], with tables, power, and Wi-Fi.
• Format: 15-minute mini-lecture → 60-minute guided coding → 15-minute Q&A/show-and-tell.
• Experience level: friendly to newcomers; helpful for practitioners who want a 2025 refresh.

First session agenda (75–90 minutes)

1. Why deep learning now? The Keras 3 ecosystem in 2025. 
2. Your first neural network in Keras (classification).
3. What’s next: where vision, timeseries, LLMs, and diffusion fit in our roadmap. 

RSVP to save your seat—bring your curiosity and a laptop. Let’s learn by building!

Discord server: https://discord.gg/WVRpVGUH

Intro to Deep Learning (Using Deep Learning with Python, 3rd Ed.)

Curious about deep learning but don’t know where to start? Join our in-person, hands-on meetup at the library as we work through François Chollet’s brand-new Deep Learning with Python, Third Edition. We’ll demystify neural nets from first principles and build real models together using Python. The book is a complete, modern rewrite that now covers Keras 3, TensorFlow, PyTorch, JAX, and today’s generative AI techniques—all taught by the creator of Keras. 

What we’ll cover (highlights from the new 3rd edition)
• Foundations: core deep learning ideas and a code-first, intuitive approach. 
• Modern tooling: Keras 3 across multiple backends, with primers on TensorFlow, PyTorch, and JAX. 
• Computer vision: image classification, image segmentation, and an intro to object detection. 
• Timeseries: practical forecasting workflows. 
• NLP & LLMs: text classification, machine translation, and large language models—including building a small GPT-like model and understanding sampling/decoding. 
• Generative images: diffusion-based image generation (the tech behind modern AI art tools). 
• Scaling & tuning: tips for faster training, multi-GPU/TPU, and model optimization. 
• Bigger & better than before: the third edition adds ~30% more content and reflects the 2024–2025 state of the art. 

Who it’s for
• Python-comfortable beginners to ML (you don’t need prior machine learning or heavy math). 

How we’ll learn
• Short explanations + live coding in Python each session.
• We’ll follow the book’s examples (with cross-references to PyTorch/JAX where helpful).
• Optional: run the official companion notebooks from the authors’ GitHub. 

Getting the book
• You can read the 3rd edition online for free on the official site, with runnable code links in each chapter. https://deeplearningwithpython.io/
• Release details: Manning lists the print/ePub/liveBook availability as September 30, 2025; Amazon lists the Kindle edition publication date as October 14, 2025. 

What to bring
• Laptop with Python 3 and a Jupyter environment. We’ll start with Keras 3 + TensorFlow; PyTorch/JAX are welcome too. (We’ll share an environment setup link at the start.) 

Location & vibe
• Where: [Library Meeting Room], with tables, power, and Wi-Fi.
• Format: 15-minute mini-lecture → 60-minute guided coding → 15-minute Q&A/show-and-tell.
• Experience level: friendly to newcomers; helpful for practitioners who want a 2025 refresh.

First session agenda (75–90 minutes)

1. Why deep learning now? The Keras 3 ecosystem in 2025. 
2. Your first neural network in Keras (classification).
3. What’s next: where vision, timeseries, LLMs, and diffusion fit in our roadmap. 

RSVP to save your seat—bring your curiosity and a laptop. Let’s learn by building!

Discord server: https://discord.gg/WVRpVGUH