Has anyone read the #FastAI "Deep learning for coders" book recently?
I got stuck at the very first chapter and it looks like I'm not alone https://github.com/fastai/course22-web/issues/42
![Running initial script (e.g. of chapter 1 of colab notebook) of the book : #hide ! [ -e /content ] && pip install -Uqq fastbook import fastbook fastbook.setup_book()_ Results in: ━━━━━━━━━━━━━━━━━━...](https://media.mstdn.social/cache/preview_cards/images/069/589/623/original/a7b7cab8e89ace06.png "Running initial script (e.g. of chapter 1 of colab notebook) of the book : #hide ! [ -e /content ] && pip install -Uqq fastbook import fastbook fastbook.setup_book()_ Results in: ━━━━━━━━━━━━━━━━━━...")
"We've now created our first deep learning neural network from scratch. And we did it in Microsoft Excel, everyone's favorite artificial intelligence tool."
- Jeremy Howard in https://www.youtube.com/watch?v=hBBOjCiFcuo&t=3862s
DeepPrep: An accelerated, scalable, and robust pipeline for neuroimaging preprocessing empowered by #deeplearning ( powered by fastSurfer, fastSurferCNN, fastCSR, Synthmorph, etc) : https://www.biorxiv.org/content/10.1101/2024.03.06.581108v1.abstract?%3Fcollection=
[Перевод] Новая R&D лаборатория (старого типа)
Самая важная новость в области ИИ последних недель, о которой вы возможно ещё не слышали. И нет, дело не в Gemini (от Google). Эрик Рис (создатель The Lean Startup) и Джереми Говард (сооснователь Kaggle, fast.ai ) объединяются, чтобы создать новый вид лаборатории искусственного интеллекта. Он называется Answer.AI . Answer.AI – это новая лаборатория, которая не ставит своей целью создание новых моделей. Вместо этого она пытается найти практическое и ценное применение существующим моделям. Их миссия заключается в том, чтобы проводить фундаментальные исследования для организаций, чтобы помочь им раскрыть возможности использования ИИ. «В Answer.AI мы не работаем над созданием AGI. Вместо этого мы заинтересованы в эффективном использовании уже существующих моделей. Выяснение того, какие практически полезные приложения могут быть построены на основе уже существующих базовых моделей, является огромной задачей, и я считаю, что ей уделяется недостаточно внимания.» – говорит Джереми Говард. Именно это меня больше всего радует – их ориентация на применение. На мой взгляд, за всей этой шумихой вокруг ИИ сегодня, скрывается реальная ценность для бизнеса, которая ждёт своего часа. Но на самом деле нет. В действительности должна существовать реальная человеческая ценность, которую нам ещё только предстоит открыть. Подобно электричеству или отвёртке, ИИ — это просто ещё один инструмент в нашем распоряжении. Только благодаря целенаправленному применению этих инструментов начинает проявляться реальная ценность. ___________________________________ Ник Роко (Nick Roco). Основатель и генеральный директор StellarStream AI, агентства по автоматизации ИИ, создающего индивидуальные инструменты ИИ для более эффективного развития бизнеса. Всё новое — это хорошо забытое старое
I finished the first chapter of the fast.ai course on Machine Learning, and I am amazed! It's incredible how easy now people without a technical background (aka me) can start diving into ML and actually doing something cool, like creating an image recognition script. By following the Jupyter Notebook code in the course I created a script to differentiate between images of birds and forests, and later on I modified it to recognize Airbus A320 vs Boeing 737 :D
I've been writing a series of articles where I run through the #fastai #deeplearning book. Except I've decided to attempt it in #rustlang with the #dfdx crate. Why not give it a read if that peaks your interest?
https://dev.to/favilo/working-through-the-fastai-book-in-rust-part-4-1ed4
Tomorrow night, learn about #nbdev. Our speaker, Kevin Bird, is a contributor to the #FastAI community, and will show us how #jupyter notebooks could transform your development. This is a free event at the #Madison Public Library (Central Branch), & is open to all skill levels. There will be free pizza 
!
https://www.meetup.com/madison-python/events/296600229/
Stick around after, and join me in pressuring Kevin to leave the bird site and join us here in the fediverse 
One week from tonight, join us to learn about #nbdev and the radically transformative power of development within #jupyter notebooks. Our speaker, Kevin Bird, is a contributor to the #FastAI community and expert on #MachineLearning and #LiterateProgramming. All skill levels, from total beginners to software experts, are welcome. The event, held at #Madison Public Library, is free and open to the public. Free pizza and beverages 
!
Just announced: "Python Development Reimagined: The Power of nbdev and Jupyter Notebooks"
Next month, come learn about nbdev, elevate #Jupyter notebooks in your development process, and make your coding life easier. This will be a great talk with something for #python newbies & pros alike.
Our speaker, Kevin Bird, is a data scientist & OSS contributor to #fastai & #nbdev
https://www.meetup.com/madison-python/events/296600229/
#Madison #MadisonWI #UWMadison
PS: Looking for sponsors for pizza & beverages. Please reach out!
Can't agree more with Rachel Thomas from #fastai https://vimeo.com/214233053
How to Learn Deep Learning (when you’re not a computer science PhD)
Coding your own AI in 2023 with fastai
by Vincent Will @wweb_dev
#CodingAI
#fastai
#webdev
#coding
#python
#deeplearning
#ai
#beginners
https://dev.to/vincenius/coding-your-own-ai-in-2023-with-fastai-2pbk
@dataschrat I like the #fastAI concept. They go end to end from the beginning. And dive deeper and deeper later on. Iterating fast is key. Setup can be painful, for me the hardest part. https://course.fast.ai/
#introduction
Dear all!
I am here on my #DataScience journey.
Studied #statistics a while ago in #munich
Working in pharma as a biostatistican for many years. Due to single parenting these days only two days a week. My sons are 12 and 16. (Their mother died 3 years ago due to cancer)
My passion is on predictive modelling. Since my personal crisis is almost over, I am back on track. This time on the #fastAI train, hoping to find some companions on my way.
Best regards,
Vielauge
Today was about starting to better understand calculus. The maths has always been my Achilles heel for the #fastai course and I can’t afford to wait until we cover those topics in my #openuniversity degree, so I’m diving in now.
Watched the beginnings of the #3blue1brown video series and started to develop a very high-level understanding.
Two key concepts: derivatives and integrals.
I’ll write a blog tomorrow summarising the extent of my understanding. #100DaysOfFastAI
@eddiekirkland add a command to export all notebooks at the end of each of your notebooks. (It's probably in the docs.) Then train yourself to always run the whole notebook once you feel you're done with your task. That way unit tests will run as well. We have a big project at AMD that uses nbdev that way. #nbdev #fastai
@victorvicpal it’s a really impressive effort. I feel like it’s the go to standard for everything that requires 3D model. For 2D Models in #medicine I still see alot of #fastai being used :)
@iamgianluca 
Hi fellow #fastai alumni !
@b0rk the tech. is not perfect yet, but you could try experimenting with using #stablediffusion Image to Image models.
Here's a notebook (run on Colab) from an ongoing #fastai course, if scroll down to the Image to Image section, there's an example of a doodle -> wolf under moonlit sky image generation.
And if you're curious, the community forums are a very welcome place. And @pcuenq could probably answer questions on the #diffusers library innerworkings.
https://github.com/fastai/diffusion-nbs/blob/master/stable_diffusion.ipynb
#Introduction I work as an escalation engineer at #Microsoft but also into #ML and #AI and working through the current #fastai course. I'm also a keen photographer and getting into #motiongraphics #vfx with #Cinema4d. The picture is from a recent trip to Joshua Tree National Park in California - and #Mastodon Peak!
