I got to interview @jeremyphoward ! 🐐🐐
When I was first trying to wrap my head around GPUs/deep learning, I spent weeks watching @fastdotai on YouTube
Crazy full circle moment to be able to give Jeremy a DGX spark
18 months ago, @karpathy set a challenge: "Can you take my 2h13m tokenizer video and translate [into] a book chapter".
We've done it! It includes prose, code & key images. It's a great way to learn this key piece of how LLMs work.
https://t.co/aSgsZz0VxO
It's a strange time to be a programmer—easier than ever to get started, but easier to let AI steer you into frustration. We've got an antidote that we've been using ourselves with 1000 preview users for the last year: "solveit"
Now you can join us.🧵
https://t.co/GLKm0woI8b
We just released a new version of fastai that's now compatible with the latest @PyTorch 2.2.
h/t @benjamin_warner
Either: "pip install -U fastai" or "conda install -c fastai fastai" to upgrade.
My @fastdotai cofounder @math_rachel has been studying immunology for the last couple of years, & she's discovered some fascinating opportunities for AI in immunology.
Her new post summarises recent research on using AI/ML to uncover immune cell networks
https://t.co/JclKG2b510
Last year, @HamelHusain (re) taught me a super power 🙏
“If you sincerely spend 3 hours everyday learning a new topic
In 6 months, you’ll be really far ahead”
This is something we learned in @fastdotai that Hamel reminded me when I expressed my imposter scare of LLMs
@jeremyphoward@fastdotai Soooo proud to be @fastdotai alumni 👨🎓 I was stuck in math previously. @fastdotai guided me into right direction. Can’t imagine only 2 months after the course, I co-founded startup with https://t.co/RJ6kWy6RP9 Alumni and won $10,000 check 🌻🚀
Second notebook "Detect Sleep States" @kaggle competition is done 🎃 it contains:
▫️ EDA & visuals of time series data
▫️ sleep representations using spectrograms
▫️ time feature engineering using @fastdotai
▫️ @wandb tracking for datasets and visuals
📍Link to notebook: https://t.co/Ev6sRCE3qx
#MadeOnZ #Datavisualization
I am so excited to see @jeremyphoward as a guest on @latentspacepod. The @fastdotai book and course helped to get me started with deep learning and also brought together an excellent community of fellow learners.
New fastai release with a handful of items+fixes from yours truly. Including adding support for bfloat16 mixed precision training.
Release has fixes from multiple new & returning contributors.
Minimum versions of Python & PyTorch were increased to 3.8 and 1.10, respectively.
As a co-leader of the Deep Learners Study Group on the @fastdotai Discord, I'm excited to announce our group is now free format!
Join us every Saturday at 12:00 UTC, regardless whether you're doing fastai Course Part 1 or 2, and share your progress!
https://t.co/M0IWLUOUD6
Good list - but another, often better way is to "consumption drive" your learning so you don't give up halfway.
Figure out the problem you want to solve, and then learn what you need to learn to solve it.
I recommend @fastdotai by @jeremyphoward
It’s cool to me how many @fastdotai alums are working on LLM stuff now. It’s a real testament to how the course didn’t just teach us specific knowledge but also a practical, problem-solving approach to ML rooted in open source (And I’ve been seeing @jeremyphoward in Discord too!)
GPU rich vs poor: Remember DAWNBench? @fastdotai blew away the leaderboard using dramatically less compute and less time
GPU abundance can actually become a weakness
https://t.co/2rl6o0Mg3W
A few months ago as part of Walk with @fastdotai, I did a presentation discussing @huggingface Accelerate. This talked about what the framework is, its layers, and how to utilize it best for your problems.
Today I've made this lecture public: https://t.co/3rebW1J5j8