Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Jeremy Howard Profile picture
@jeremyphoward
Jul 28, 2022 • 10 tweets • 8 min read • Read on X
Scrolly
Our biggest launch in years: nbdev2, now boosted with the power of @quarto_pub!

Use @ProjectJupyter to build reliable and delightful software fast. A single notebook creates a python module, tests, @github Actions CI, @pypi/@anacondainc packages, & more
fast.ai/2022/07/28/nbd…
What can you create with #nbdev with @quarto_pub? Well, for starters, check out our beautiful new website.

Created with nbdev, of course!

nbdev.fast.ai
That website is generated automatically from the notebooks in this repo.

Take a look around -- all the things you'd hope to see in a high-quality project are there. It's all done for you by nbdev. For instance, see the nice README? Built from a notebook!
github.com/fastai/nbdev/
nbdev v1 is already recommended by experts, and v2 is a big step up again.

"From my point of view it is close to a Pareto improvement over traditional Python library development." Thankyou @erikgaas 😀
Here's an example of the beautiful and useful docs that are auto-generated by nbdev+Quarto.
nbdev.fast.ai/merge.html
Here's an example of an exported function in a notebook cell. This is automatically added to the python module, and the documentation you see on the right is auto-generated.
nbdev.fast.ai/merge.html#nbd…
Every time we update a notebook to change the docs, library, or tests, everything is checked by @github Actions automatically
Here's the @pypi pip installer that's auto-generated. See the description? That's created for you from the notebook you use for your documentation homepage (just like the README, and the description for your conda package)
pypi.org/project/nbdev/
I've barely scratched the surface in this brief tweet thread! For much more information, take a look at the blog post authored with @HamelHusain
fast.ai/2022/07/28/nbd…
This launch wouldn't have been possible without some amazing people. I'd especially like to highlight Hamel & @wasimlorgat, who made nbdev2 a far better product than it would have been without them, JJ Allaire @fly_upside_down & the @quarto_pub team, & the @fastdotai community

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Jeremy Howard

Jeremy Howard Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @jeremyphoward

Jeremy Howard Profile picture
Mar 29
I'm glad @levelsio checked this, but sad our contrib has been erased by later big tech co's. Alec Radford said ULMFiT inspired GPT. ULMFiT's first demo predated BERT.

Today's 3-stage LLM approach of general corpus pretraining and 2 stages of fine-tuning was pioneered by ULMFiT.
There have been many other important contributions, including attention (Bahdanau et al), transformers, RLHF, etc.

But before all this, basically everyone in NLP assumed that each new domain needed a new model. ULMFiT showed that a large pretrained model was actually the key.
I got push-back from pretty much everyone about this. My claim that fine-tuning that model was the critical step to achieving success in NLP was not something people were ready to hear at that time.

I gave many talks trying to convince academics to pursue this direction.
Read 5 tweets
Jeremy Howard Profile picture
Mar 18
Announcing fasttransform: a Python lib that makes data transformations reversible/extensible. No more writing inverse functions to see what your model sees. Debug pipelines by actually looking at your data.

Built on multi-dispatch. Work w/ @R_Dimm
fast.ai/posts/2025-02-…
We took the `Transform` class out of fastcore, replaced the custom type dispatch system with @ikwess's plum-dispatch, mixed it all together, and voila: fasttransform! :D

To learn about fasttransform, check out our detailed blog post.
fast.ai/posts/2025-02-…
"Manual inspection of data has probably the highest value-to-prestige ratio of any activity in machine learning." --@gdb

Yet we often skip it because it's painful. How do you inspect what your model sees after normalization, resizing & other transforms?
Read 10 tweets
Jeremy Howard Profile picture
Feb 5
Wow, actual grown men are still doing the "I asked the LLM about itself and it said" thing.

In 2025.

Folks, LLMs don't know anything about how they themselves are built or deployed, unless they've been explicitly programmed with that information (which they almost never are). Image
If you are using a Chinese cloud based service hosted in China, then your data will be sent to a server in China.

Because that's how computers work.

It's not a conspiracy.
I hadn't heard of this guy before, but apparently he'll teach you how to learn AI.

Cool. Image
Read 4 tweets
Jeremy Howard Profile picture
Jan 25
I've recently been surprised to discover that a few of my friends are choosing to use nicotine to help them with focus, even though they are not ex-smokers.

I decided to look into it, and it turns out that there are documented health benefits of nicotine for some people. 🧵
I specifically looked into nicotine for ADHD, since, at least among children, ADHD and giftedness go hand in hand statistically (which would apply in adulthood too), and because focus was mention as an area where nicotine can be helpful.
There is a great overview below. But "Very surprisingly, there are… no further… studies.

Research into active ingredients… is expensive.

In addition, nicotine has a very poor image… which impairs its marketability"
adxs.org/en/page/192/ni…
Read 12 tweets
Jeremy Howard Profile picture
Dec 19, 2024
I'll get straight to the point.

We trained 2 new models. Like BERT, but modern. ModernBERT.

Not some hypey GenAI thing, but a proper workhorse model, for retrieval, classification, etc. Real practical stuff.

It's much faster, more accurate, longer context, and more useful. 🧵 Image
ModernBERT is available as a slot-in replacement for any BERT-like model, with both 139M param and 395M param sizes.

It has a 8192 sequence length, is extremely efficient, is uniquely great at analyzing code, and much more. Read this for details:
huggingface.co/blog/modernbert
Seven months ago, @bclavie kicked things off, and soon @benjamin_warner & @antoine_chaffin joined him as project co-leads. I don't think anyone quite knew what we were getting in to…

It turns out that training a new, SoTA model from scratch is actually pretty hard. Who knew? 🤷 Image
Read 23 tweets
Jeremy Howard Profile picture
Nov 16, 2024
Oh wow. This is gonna be super tricky to figure out what to do now.

There isn’t any easy automated way to install a full deep learning stack from scratch after this change afaict.
I wonder if the @PyTorch analysis behind this is mistaken. I suspect most of the pypi installs they’re seeing are from CI and similar. Conda installs are the standard for end user installation of PyTorch afaik
@PyTorch Conda aggressively caches installs so looking at relative download numbers won’t give a great sense of real usage.
Read 7 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(