Jeremy Howard @jeremyphoward Twitter profile

Pinned Tweet

Jeremy Howard

5 months

Today @ericries (creator of Lean Startup & LTSE) & I ( /Kaggle) are launching a new kind of R&D lab: . We're backed by $10m of funding from @DecibelVC . For-profit R&D labs are rare today, but have an amazing history... 🧵

203

332

3K

Last Seen Profiles

@ilovejohnkimble

@InTheZoneStats_

@ahibuk5

@monaghan_lizy

@yahya77alshiba2

@fly_BUR

@excallingg

@MichelleDLesley

@kstnz

@fch_21

@blonde_bandit69

@tulilo83292

@OmniPayGroup

@paul0m0nter

@FosteringHopeOH

@Johnny_MacKay28

@GeorgiaPower

@gigafactories

@SindreEspe

@imMissBeliever

@wrttn_word

@JordanNolin2

@SergioSarmiento

@subinium

@Saudi_FETP

@coachbom

@KevinHartmannC

@hmsahawaii

@syaaaa_lah

@motykamilosz

@AlvarezMaynez

@ashidmunkh3

@AamirBm282965

@kittischmitti

@WhaleBetting

@canbequit

Jeremy Howard

@jeremyphoward

4 years

So do these non-medical masks we're all meant to be wearing actually do anything? Well yes, yes they almost certainly do. Here's a thread about the current science. Let's start by explaining this striking picture of a researcher speaking in a laser scattering chamber. 1/n

191

10K

16K

Jeremy Howard

@jeremyphoward

4 years

The Czech Republic went from zero mask usage to 100% in 10 days, and in the process they halted the growth of new covid-19 cases. How? They made their own! They didn't need government help; they did it themselves. It's time for #masks4all . See why: 1/

We need #masks4all

Update: We now have a website - www.Masks4All.co . We've compiled research, expert voices, and DIY recipes, all in one regularly updated place.In the Czech R...

www.youtube.com

344

5K

9K

Jeremy Howard

@jeremyphoward

9 months

Now that ChatGPT has rolled out custom instructions to most users, try out this instruction -- it makes GPT 4 far more accurate for me: (Concat the rest of this 🧵 together and put in your custom instruction section)

152

775

7K

Jeremy Howard

@jeremyphoward

8 months

I just uploaded a 90 minute tutorial, which is designed to be the one place I point coders at when they ask "hey, tell me everything I need to know about LLMs!" It starts at the basics: the 3-step pre-training / fine-tuning / classifier ULMFiT approach used in all modern LLMs.

88

1K

7K

Jeremy Howard

@jeremyphoward

4 years

As data scientists, we (with @math_rachel ) have done our best to look at the data around covid-19, and what it means to you and your community. Our view: it is appropriate to be very concerned, and significantly change your lifestyle, right now. 1/13

119

4K

6K

Jeremy Howard

@jeremyphoward

7 months

Here, in full directly on Twitter, is "A Hackers' Guide to Language Models". This 90 minute tutorial is designed to be the one place I point coders at when they ask "hey, tell me everything I need to know about LLMs!" It covers both @OpenAI models and open source ones in depth.

83

1K

6K

Jeremy Howard

@jeremyphoward

3 months

Also unfortunately not satire

𝕝𝕚𝕘𝕙𝕥𝕟𝕚𝕟𝕘𝕗𝕚𝕣𝕖 𝕤𝕥𝕠𝕜𝕖𝕤

@spooky_stokes

3 months

unfortunately not satire

128

52

2K

61

144

6K

Jeremy Howard

@jeremyphoward

4 years

New analysis. Texas re-opened early, without requiring masks. Things are looking bad. NB: chart shows hospitalizations, not cases, so this is not a side-effect of increased testing.

265

3K

5K

Jeremy Howard

@jeremyphoward

6 months

OK everyone's asking me for my take on the OpenAI stuff, so here it is. I have a strong feeling about what's going on, but no internal info so this is just me talking. The first point to make is that the Dev Day was (IMO) an absolute embarrassment.

182

578

5K

Jeremy Howard

@jeremyphoward

2 years

After 2 years, Practical Deep Learning for Coders v5 is finally ready! 🎊 This is a from-scratch rewrite of our most popular course. It has a focus on interactive explorations, & covers @PyTorch , @huggingface , DeBERTa, ConvNeXt, @Gradio & other goodies 🧵

Practical Deep Learning for Coders - Practical Deep Learning

A free course designed for people with some coding experience, who want to learn how to apply deep learning and machine learning to practical problems.

course.fast.ai

116

1K

5K

Jeremy Howard

@jeremyphoward

1 year

Our new course, "From Deep Learning Foundations to Stable Diffusion", is finally done after 8 months of work!!! With >30 hours of video content (all free, no ads!), you'll learn how to create and train a Stable Diffusion model starting from pure Python 🧵

112

1K

5K

Jeremy Howard

@jeremyphoward

1 year

There's a new programming language in town - it's Mojo! I'm more than a little excited about it. It's Python, but with none of Python's problems. You can write code as fast as C, and deploy small standalone applications like C. My post is below, and a 🧵

fast.ai - Mojo may be the biggest programming language advance in decades

Mojo is a new programming language, based on Python, which fixes Python’s performance and deployment problems.

www.fast.ai

107

936

5K

Jeremy Howard

@jeremyphoward

2 years

I led the team that studied mask efficacy in early 2020 and published our results in the Proceedings of the National Academy of Science. I spent three months earlier this year revisiting this topic, and today I'm publishing my notes and links here:

155

2K

5K

Jeremy Howard

@jeremyphoward

4 years

This is huge. The creator of the YOLO algorithms, which (along with SSD) set much of the path of modern object detection, has stopped doing any computer vision research due to ethical concerns. I've never seen anything quite like this before.

Joseph Redmon

@pjreddie

4 years

I stopped doing CV research because I saw the impact my work was having. I loved the work but the military applications and privacy concerns eventually became impossible to ignore.

77

918

3K

47

2K

5K

Jeremy Howard

@jeremyphoward

8 months

I was chatting on Discord with an anon account ,who had just come up with a really smart new approach to model merging and had some encouraging results to show. But then he had to go. Turns out he's in 10th grade and was in trouble because he wasn't focussing on English class.

63

343

4K

Jeremy Howard

@jeremyphoward

15 days

In Australia we have a computational thinking competition for kids each year called Bebras. You get a certificate if you get all of them right. Here's one of the "easy" level questions for children in grade 3. GPT 4 got it wrong.

92

251

4K

Jeremy Howard

@jeremyphoward

24 days

"transformers or diffusion" Ummm.... who's gonna tell him that that makes no sense at all? (Transformers is an architecture, and it's quite common to use it *with* diffusion.)

Shubham Saboo

@Saboo_Shubham_

25 days

Elon Musk’s lit reply to a reporter 🔥 REP: Who do you think is going to win The White House in 2024? Elon: Who do you think will win the White House in 2032? Which type of AI? Transformers or Diffusion?

58

29

324

189

170

4K

Jeremy Howard

@jeremyphoward

4 years

Today is ’s biggest day in our four year history. We are releasing our new course, fastai v2, fastcore, fastscript, fastgpu, and the official launch of the book. 1/

93

990

4K

Jeremy Howard

@jeremyphoward

2 months

Today, with @Tim_Dettmers , @huggingface , & @mobius_labs , we're releasing FSDP/QLoRA, a new project that lets you efficiently train very large (70b) models on a home computer with consumer gaming GPUs. 1/🧵

Answer.AI - You can now train a 70b language model at home

We’re releasing an open source system, based on FSDP and QLoRA, that can train a 70b model on two 24GB GPUs.

www.answer.ai

87

685

4K

Jeremy Howard

@jeremyphoward

3 years

Our paper, "An evidence review of face masks against COVID-19", written by a cross-disciplinary team of 19 international experts, was published in the Proceedings of the National Academy of Sciences today. No time for the paper? Then read this thread!

An evidence review of face masks against COVID-19 | Proceedings of the National Academy of Sciences

The science around the use of masks by the public to impede COVID-19 transmission is advancing rapidly. In this narrative review, we develop an ana...

www.pnas.org

111

2K

4K

Jeremy Howard

@jeremyphoward

4 years

I completed my 1st data science project ~30 years ago. Since then I've been continuously developing a questionnaire I use for all new data projects, to ensure the right info is available from the start. I'm sharing it publicly today for the first time.

52

970

3K

Jeremy Howard

@jeremyphoward

3 years

Wow. @OpenAI has trained a model that can create new images - you just describe them in English. 🤯

43

853

3K

Jeremy Howard

@jeremyphoward

2 years

Big news: we're launching a new course in <4 weeks. "From Deep Learning Foundations to Stable Diffusion". Bigger news: for this course, we're teaming up with @StabilityAI ! AFAIK, this is the 1st course that covers every method used in Stable Diffusion.

41

655

3K

Jeremy Howard

@jeremyphoward

5 years

This single line of code solves a years-old bug that has caused problems for thousands of people. No one had figured it out before. It took @StasBekman , one of the best programmers I know of, weeks of work to develop. This is why we don't measure productivity with lines of code

26

1K

3K

Jeremy Howard

@jeremyphoward

4 years

In our book we always show code equivalents for any math notation. So often, the Python & @PyTorch code is clearer and simpler. And you can actually execute it and try it! For instance, LogSumExp:

55

500

3K

Jeremy Howard

@jeremyphoward

2 years

I got a special surprise for you all... We just released the first 5.5 hours of our new course "From Deep Learning Foundations to Stable Diffusion", for free!

55

595

3K

Jeremy Howard

@jeremyphoward

3 years

I'm really excited to say that I'm sending this tweet from Australia - our new home!😃 I learned lots in SF and had a great experience, but we've decided that long-term the US isn't really the right place for us to raise a family. @fastdotai will keep going, with an Aussie base

154

70

3K

Jeremy Howard

@jeremyphoward

4 years

We've made a draft of our deep learning book available as @ProjectJupyter notebooks. Important: This is for personal use, and not licensed for any redistribution or change of format or medium. Eg. don't make a PDF and share it, please!

61

894

3K

Jeremy Howard

@jeremyphoward

4 years

This is one of the most amazing and hopeful things I've read. How the Czech Republic, through community effort, got to 100% mask usage in 10 days, nearly all thru home DIY effort. They only have 2 deaths and have no growth in daily new covid-19 cases.

100

1K

3K

Jeremy Howard

@jeremyphoward

2 years

I am over the moon to announce: 1) I'm now a professor at University of Queensland (UQ), the top institute in my home state! 2) I'll be teaching a brand new deep learning course at UQ from April, which will form the basis of a new @fastdotai course! 🧵

133

184

3K

Jeremy Howard

@jeremyphoward

1 year

Don't listen to me. I don't understand language model fine tuning. I'm merely the 1st author of the paper "Universal Language Model Fine Tuning", which explained 5 years ago how to fine tune universal language models.

53

207

3K

Jeremy Howard

@jeremyphoward

5 years

Wow wow wow! @GoogleColab now supports the fastai library directly :D That means you can run all the lessons at , and do your own research and development with fastai, all for free. That's... that's just amazing. To get started:

45

700

3K

Jeremy Howard

@jeremyphoward

4 years

I'm happy this little-known data scientist helped make a difference! 😊

74

254

3K

Jeremy Howard

@jeremyphoward

3 months

I used to find writing CUDA code rather terrifying. But then I discovered a couple of tricks that actually make it quite accessible. In this video I introduce CUDA in a way that will be accessible to Python folks, & I even show how to do it all in Colab!

Getting Started With CUDA for Python Programmers

I used to find writing CUDA code rather terrifying. But then I discovered a couple of tricks that actually make it quite accessible. In this video I introduc...

www.youtube.com

21

373

3K

Jeremy Howard

@jeremyphoward

1 year

There's a lot of folks under the misunderstanding that it's now possible to run a 30B param LLM in <6GB, based on this GitHub discussion. This is not the case. Understanding why gives us a chance to learn a lot of interesting stuff! 🧵

30B model now needs only 5.8GB of RAM? How? · ggerganov llama.cpp · Discussion #638

(Edit: apologies, I should have clarified initially I'm running on Linux OS. I didn't realize it might not be obvious from the screenshot alone for a non-Linux users.All tests are done on U...

github.com

24

419

3K

Jeremy Howard

@jeremyphoward

7 months

If you're like me and find it easier to read math than code, and you have access to @OpenAI GPT 4V, try pasting a image of an equation you wanna understand in there. It might just blow your mind.

60

335

3K

Jeremy Howard

@jeremyphoward

3 months

I used to find writing CUDA code rather terrifying. But then I discovered a couple of tricks that actually make it quite accessible. In this video I introduce CUDA in a way that will be accessible to Python programmers, and I even show how to do it all in @GoogleColab !

36

408

3K

Jeremy Howard

@jeremyphoward

3 years

I'm the lead author of the first and largest academic evidence review of masks, published in the Proceedings of the National Academy of Sciences, with 141 references. The claim that evidence doesn't support use of masks in kids in classrooms couldn't be more wrong. 1/🧵

Dr. Nick Coatsworth

@nick_coatsworth

3 years

No masks on kids in classrooms. Period. Evidence doesn’t support it. Vaccinated adults the best protection for kids. Zero COVID academics should leave our kids alone. #backtoschool #auspol #covid19nsw #covid19vic

442

288

1K

94

1K

3K

Jeremy Howard

@jeremyphoward

5 years

Practical Deep Learning for Coders, 2019 edition, is now available. With a shiny new video player with searchable transcripts. This course is 100% new material, including some new techniques and results never previously published.

32

895

3K

Jeremy Howard

@jeremyphoward

3 months

GitHub banned their own co-founder, because something "went wrong with our automations". Please folks, do NOT use automated decision making tools to make decisions that can harm people.

defunkt

@defunkt

3 months

Banned from GitHub without any explanation. Guess I'm moving all my code to BitBucket.

552

829

9K

81

244

2K

Jeremy Howard

@jeremyphoward

5 years

"New State of the Art AI Optimizer: Rectified Adam (RAdam). Improve your AI accuracy instantly versus Adam, & why it works" It's been a long time since we've seen a new optimizer reliably beat the old favorites; this looks like a very encouraging approach!

New State of the Art AI Optimizer: Rectified Adam (RAdam).

A new paper by Liu, Jian, He et al introduces RAdam, or “Rectified Adam”. It’s a new variation of the classic Adam optimizer that provides…

link.medium.com

19

833

2K

Jeremy Howard

@jeremyphoward

2 years

TIL it's this easy to plot math functions in Python, thanks to @SymPy . No more mucking about manually with np.linspace and matplotlib and friends...

29

319

2K

Jeremy Howard

@jeremyphoward

6 months

Bitcoin-turned-AI bros updating their profiles right now. ❤️ b̵i̵t̵c̵o̵i̵n̵ n̵f̵t̵ w̵e̵b̵3̵ A̵I̵ g̵e̵n̵e̵r̵a̵t̵i̵v̵e̵ RL/q-learning

81

202

2K

Jeremy Howard

@jeremyphoward

6 years

Proud to be launching "Introduction to Machine Learning for Coders" today. If you're looking for a modern, code-first, opinionated introduction to machine learning, I think you might like this.

25

750

2K

Jeremy Howard

@jeremyphoward

4 years

We're launching `fastpages`, a platform which allows you to host a blog for free, with no ads. You can blog with @ProjectJupyter notebooks, @office Word, directly from @github 's markdown editor, etc. Nothing to install, & setup is automated!

Introducing fastpages

An easy to use blogging platform with extra features for Jupyter Notebooks.

fastpages.fast.ai

41

644

2K

Jeremy Howard

@jeremyphoward

10 months

I've spent the last few months interviewing >60 experts in law, economics, AI, alignment, etc, on the impacts of AI, and safety interventions. Today I'm publishing my first article, showing regulation designed to increase AI safety may backfire badly!

fast.ai - AI Safety and the Age of Dislightenment

Model licensing & surveillance will likely be counterproductive by concentrating power in unsustainable ways

www.fast.ai

78

563

2K

Jeremy Howard

@jeremyphoward

2 years

I've just published over 20 hours of tutorials and live coding showing how to: install python the right way; set up a terminal; write shell scripts; use vim; use a remote Jupyter server; use git, github, tmux, and ssh; use the python debugger; and more! 🧵

fast.ai live coding & tutorials

Over 20 hours of tutorials and live coding showing how to: install python the right way; set up a terminal; write shell scripts; use vim; use a remote Jupyte...

www.youtube.com

32

427

2K

Jeremy Howard

@jeremyphoward

6 years

Learn: - Intro Machine Learning - Practical Deep Learning - Cutting Edge Deep Learning - Computational Linear Algebra

GitHub - fastai/numerical-linear-algebra: Free online textbook of Jupyter notebooks for fast.ai...

Free online textbook of Jupyter notebooks for fast.ai Computational Linear Algebra course - fastai/numerical-linear-algebra

github.com

23

644

2K

Jeremy Howard

@jeremyphoward

4 months

If you're a Python programmer looking to get started with CUDA, this weekend I'll be doing a free 1 hour tutorial on the absolute basics. Thanks to @neurosp1ke , @marksaroufim , and @ThomasViehmann for hosting this on the CUDA MODE server. :D Click here:

Discord - A New Way to Chat with Friends & Communities

Discord is the easiest way to communicate over voice, video, and text. Chat, hang out, and stay close with your friends and communities.

discord.com

31

283

2K

Jeremy Howard

@jeremyphoward

4 years

What's happening here? Does this masks thing really make sense, scientifically? Yes, it does! Here's 34 scientific papers that all point to the same thing: masks (even home made ones) really work to dramatically reduce the spread of covid-19.

40

1K

2K

Jeremy Howard

@jeremyphoward

7 months

Wow somehow I've been completely erased from the history of Kaggle. For the record, the US company that was funded, Kaggle Inc, was owned equally by Anthony Goldbloom and I with a minority holding by Nick Gruen. Ben Hamner didn't join until later.

91

135

2K

Jeremy Howard

@jeremyphoward

6 years

Every week I get asked "what's 's business model?" Our business model is: we spend our personal money to do stuff that helps people use deep learning. That's it. We're totally independent. We don't even take grant money.

fast.ai - fast.ai—Making neural nets uncool again

www.fast.ai

54

327

2K

Jeremy Howard

@jeremyphoward

7 months

If you're like me and find it easier to read *code* than *math*, and you have access to @OpenAI GPT 4V (or use @bing or @google Bard), try pasting a image of an equation you wanna understand in there. It might just blow your mind. 1/🧵

66

260

2K

Jeremy Howard

@jeremyphoward

1 year

I remember well when Google was first released. I felt confident that it would become a critical tool that we'd all rely on. I haven't felt that way again, until now: ChatGPT. In just a few days I've gotten so much out of it.

38

160

2K

Jeremy Howard

@jeremyphoward

4 years

The picture above shows speech droplets, which are believed to be the key transmission vector for COVID-19, with vs without a paper towel face cover. As you see, nearly all droplets are blocked. Here's a video with a cloth covering, from this NEJM paper:

16

669

2K

Jeremy Howard

@jeremyphoward

1 year

Lots of folks misunderstanding how GPT 3.5 & 4 work, such as in the take below. "Predict text" is the *pre-training* objective, but it's not what's used for the final model. The final model's objective is, in fact, very similar to adversarial loss (i.e similar to a GAN).

Eliezer Yudkowsky ⏹️

@ESYudkowsky

1 year

GPTs are not GANs. They aren't trained to *talk like a human*. They're trained to *predict all the text on the Internet*. Think about that for a second. Imagine yourself in a box, predicting those next words. Is that a task where performance caps out at human intelligence?

118

115

1K

40

228

2K

Jeremy Howard

@jeremyphoward

4 years

We've just completed a 19-author analysis of the effectiveness of mask wearing, with 84 references. To explain what the science shows, I teamed up with the wonderful Prof @trishgreenhalgh CBE, who just led a British Medical Journal study on this. (thread)

118

1K

2K

Jeremy Howard

@jeremyphoward

16 days

Today at @answerdotai we've got something new for you: FSDP/QDoRA. We've tested it with @AIatMeta Llama3 and the results blow away anything we've seen before. I believe that this combination is likely to create better task-specific models than anything else at any cost. 🧵

38

293

2K

Jeremy Howard

@jeremyphoward

5 years

Practical Deep Learning for Coders, 2019 edition, will be released tomorrow. It's looking amazing.

55

319

2K

Jeremy Howard

@jeremyphoward

5 years

Learn: - Practical Deep Learning - Deep Learning Foundations - Computational Linear Algebra - Intro Machine Learning

42

446

2K

Jeremy Howard

@jeremyphoward

8 months

It looks like @johnowhitaker & I may have found something crazy: LLMs can nearly perfectly memorise from just 1-2 examples! We're written up a post explaining what we've seen, and why we think rapid memorization fits the pattern. Summary 🧵 follows.

fast.ai - Can LLMs learn from a single example?

We’ve noticed an unusual training pattern in fine-tuning LLMs. At first we thought it’s a bug, but now we think it shows LLMs can learn effectively from a single example.

www.fast.ai

51

307

2K

Jeremy Howard

@jeremyphoward

1 year

The @karpathy approach to LLMops is the correct way

23

82

2K

Jeremy Howard

@jeremyphoward

6 months

Screw this shitty credentialist arrogant take. Helen Toner is one of the most brilliant people I've ever chatted to. I'm not sure I've ever dealt with anyone else who was so informed and thoughtful on AI ethics, policy, and safety. She's a brilliant choice for the @OpenAI board

100

74

2K

Jeremy Howard

@jeremyphoward

3 months

The actual reason that it's oriented at 3 o'clock and goes anti clockwise is because that's the correct orientation. If you're gonna criticise someone for promoting learning, you should probably get the basics right yourself first.

15

8

2K

Jeremy Howard

@jeremyphoward

9 months

I was curious about what @OpenAI GPT 3.5 would do if I used `function_call` to force it to call a function, but passed in a prompt that didn't make sense for that function. As it turns out, the model found an answer that absolutely makes sense...

54

125

2K

Jeremy Howard

@jeremyphoward

6 months

I could barely watch the keynote. It was just another bland corp-speak bunch of product updates. For those researchers I know that were involved from the beginning, this must have felt nausea-inducing. The plan was AGI, lifting society to a new level. We got Laundry Buddy.

49

62

2K

Jeremy Howard

@jeremyphoward

4 years

We have a *lot* of really great news to share with you regarding our forthcoming book, "Deep Learning for Coders", thanks to the amazing @OReillyMedia content team. 1: We're expanding it from 350 pages to ~500p, with lots of extra content...

Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD

www.amazon.com

28

299

2K

Jeremy Howard

@jeremyphoward

4 years

So if masks look so good on paper, do they work in practice? They sure seem to! In countries with universal mask-wearing, per-capita coronavirus mortality increased by just 5.4% each week compared with 48% each week in remaining countries

18

757

2K

Jeremy Howard

@jeremyphoward

5 months

“Nobody called LLMs” totally misunderstands how AI development happens. LLMs didn’t spontaneously appear for no reason. I knew they would work, I demonstrated them working in 2017, and I said at that time I think they’ll be a big deal.

Eliezer Yudkowsky ⏹️

@ESYudkowsky

5 months

Nobody called LLMs, as far as I know. One guy at OpenAI built GPT-1 while the others were working on deep RL, OpenAI Gym and such. Technology in general and AI in particular is not that predictable! People thought at one point that defeating human grandmasters at chess would…

36

12

249

46

113

2K

Jeremy Howard

@jeremyphoward

5 years

"Data Visualization: A practical introduction" A stunning, beautiful, carefully researched, free, online book by @kjhealy

6

463

2K

Jeremy Howard

@jeremyphoward

2 months

I'd given up using ChatGPT for all but the most basic tasks -- I just wasn't getting answers that were good enough to be of practical use to me. But Claude 3 Opus is being genuinely useful, and it's making me use LLM chat again. Thanks @AnthropicAI !

58

81

2K

Jeremy Howard

@jeremyphoward

4 years

A Cambridge University study showed that lots of household materials are nearly as good as surgical masks, and you can use 2 layers of material to make them even better. It doesn't need to be perfect - it just needs to reduce viral load so your immune system can kick in!

30

761

1K

Jeremy Howard

@jeremyphoward

4 years

I did a deep-dive into @GitHub Pages, and found it's possible to create a *really* easy way to host your own blog: no code, no terminal, no template syntax. I made "fast_template" to pull this together, & a guide showing beginners how to get blogging

37

361

1K

Jeremy Howard

@jeremyphoward

5 years

"PyTorch is becoming the deep learning framework of choice at Microsoft" - Big statement from @Microsoft at @PyTorch devcon

18

345

1K

Jeremy Howard

@jeremyphoward

4 years

Did you know that if you open a PDF in Microsoft Word, it gets converted into a fully editable, properly formatted word processing document? It's amazingly good.

53

301

1K

Jeremy Howard

@jeremyphoward

4 years

This library is totally amazing. I suspect that everyone building neural nets is going to want this in their life.

Terence Parr

@the_antlr_guy

4 years

Introducing TensorSensor: a lib that clarifies exceptions by augmenting messages + visualizing Python code to indicate shape of tensor variables; works with Tensorflow, PyTorch, Numpy, and higher-level libraries like Keras and fastai. Article:

29

559

2K

11

304

1K

Jeremy Howard

@jeremyphoward

4 years

I've been working with @GuggerSylvain for a long time to create something that combines the best of @ProjectJupyter Notebooks with the best of traditional software development approaches. It's called nbdev. We're releasing it today as open source. 1/

31

473

1K

Jeremy Howard

@jeremyphoward

9 months

Yay thank you @NVIDIAAI ! 😄 3 x A6000 48GB GPUs. Successfully installed into my @LambdaAPI box. OK... what shall I train first?

115

39

1K

Jeremy Howard

@jeremyphoward

3 months

Folks, you don't need anything special to have python automatically drop you into the debugger if there's an exception -- this capability is already built in! Here's a complete demonstration of how to do it, including an alias you might find pretty handy:

20

134

1K

Jeremy Howard

@jeremyphoward

6 years

After 2 years of development, we've just launched fastai v1, the first deep learning library with a simple consistent API across vision, text, tabular, and collaborative filtering data. Built on the wonderful @PyTorch v1 (preview released today)

20

491

1K

Jeremy Howard

@jeremyphoward

4 years

Never thought my worlds of deep learning and masks would collide in a single video!

NursingNotes

@NursingNotesUK

4 years

A paramedic has found an innovative way to communicate with hard of hearing patients while wearing a face mask.

508

31K

63K

13

303

1K

Jeremy Howard

@jeremyphoward

3 years

So cool - an automated pothole finder! This could be a great way to tell your local council about road problems...

Andrea 🤌🏾 Ranieri

@4ndr3aR

3 years

OH MY GOD! This one turned out really well! (Trained with @fastdotai btw) #DeepLearning #segmentation

128

708

4K

36

211

1K

Jeremy Howard

@jeremyphoward

5 years

I believe this is the first time a dataset like this has been released - this is really exciting!

Matthew Lungren MD MPH

@mattlungrenMD

5 years

🤖📢‼️ML Healthcare Researchers ‼️📢🤖 BIG medical data release thanks to @alistairewj @MIT @BIDMChealth MIMIC 2.0 >375k chest Xrays + 𝙁𝙍𝙀𝙀 𝙏𝙀𝙓𝙏 𝙍𝘼𝘿𝙄𝙊𝙇𝙊𝙂𝙔 𝙍𝙀𝙋𝙊𝙍𝙏𝙎 Link: cc: @jeremyphoward @jeremy_irvin16

21

382

1K

7

362

1K

Jeremy Howard

@jeremyphoward

5 years

Really proud to share "What is torch.nn, really?", which takes you from a neural net written from scratch, refactored step by step using all the key concepts in `torch.nn`. If you want to really understand how neural nets work in @PyTorch , start here!

11

348

1K

Jeremy Howard

@jeremyphoward

5 months

Early reports from people using this are that it's the real deal. Strong coding. Good multilingual. Consistent over long contexts. Looks to be yet another brilliant Chinese model. They now cover the spectrum of popular sizes: Qwen (7b & 14b), Yi (34b), and now Deepseek (67b).

anton

@abacaj

5 months

DeepSeek 67B looks incredibly strong on benchmarks (like a llama-2 that can also write code)

17

104

826

21

141

1K

Jeremy Howard

@jeremyphoward

7 years

After months of work, it's done! Presenting the "Deep Learning For Coders" MOOC 18 hours of lessons, all free

Practical Deep Learning for Coders - Practical Deep Learning

A free course designed for people with some coding experience, who want to learn how to apply deep learning and machine learning to practical problems.

course.fast.ai

26

619

1K

Jeremy Howard

@jeremyphoward

1 year

Wow big call from @ylecun . I wonder who or what is stopping this from happening.

47

85

1K

Jeremy Howard

@jeremyphoward

5 years

fastai has now been updated for the release version of @pytorch v1. You can install cuda, cudnn, pytorch, torchvision, and fastai with a single line. In fact, here's a pic showing all the steps necessary to install, download data, and create and fine-tune a pretrained model :)

16

331

1K

Jeremy Howard

@jeremyphoward

6 years

Oh my. This is fascinating and amazing... A totally practical approach to building a new kind of spell checker, just invented in a forum post! :)

21

349

1K

Jeremy Howard

@jeremyphoward

2 years

Are you ready to embark on a deep learning journey? I've just released over 6 hours of videos and the first in a series of notebooks showing the thought process of how I got to #1 in a current Kaggle comp. Follow this 🧵 for updates on the journey!

15

243

1K

Jeremy Howard

@jeremyphoward

5 years

Oh, nothing to see here really, just plots, in fastai, ... IN THE CONSOLE!!! Coming in the next release. :)

43

171

1K

Jeremy Howard

@jeremyphoward

5 months

I'm very happy to announce that @fastdotai is a founding member of the new AI Alliance, a consortium dedicated to the open and collaborative development and distribution of AI.

21

138

1K

Jeremy Howard

@jeremyphoward

8 months

I'm doing a 1-hour recorded talk "A hackers guide to open source LLMs" for a conference. I want to show a bit about what LLMs are, how to use them, and get folks started on OS LLMs. What things would you be interested in me covering? (The talk will be shared publicly.)

146

88

1K

Jeremy Howard

@jeremyphoward

1 year

"Any model made available in the EU, without first passing extensive, and expensive, licensing, would subject companies to massive fines of the greater of €20,000,000 or 4% of worldwide revenue. Opensource developers, and hosting services such as GitHub... would be liable"

Technomancers_ai

@technomancers

1 year

@MeetThePress @ericschmidt Fear reaction to what the EU is about to do.

32

68

322

102

234

1K

Jeremy Howard

@jeremyphoward

3 years

The Google Ethical AI team is now purged of all its management. This was the group that previously acted as an internal brake on Google building unethical AI tools. It's now been gutted, in a very methodical fashion.

MMitchell

@mmitchell_ai

3 years

I'm fired.

839

2K

9K

31

490

1K

Jeremy Howard

@jeremyphoward

10 months

For folks looking to learn NLP, the slides for Stanford's excellent cs224n courses are all available here (scroll to the bottom for the latest version):

16

246

1K

Jeremy Howard

@jeremyphoward

8 months

There's an amazingly convenient way to install the *full* NVIDIA CUDA dev stack on Linux, that I've never seen mentioned before. It's all done with conda! I just tried it and it worked perfectly.🧵

40

216

1K

Jeremy Howard

@jeremyphoward

5 years

Thank you everyone for your patience - our new deep learning MOOC is here! Includes 5 lessons diving into the foundations with @PyTorch , and 2 lessons (co-taught with @clattner_llvm ) building a new deep learning library in #swift

21

375

1K

Jeremy Howard

@jeremyphoward

1 year

I'm seeing a lot of people confused about this - asking: what exactly is the problem here? That's a great question! Let's use this as a learning opportunity and dig in. 🧵

Aidan Gomez

@aidangomez

1 year

my god..

107

46

959

65

177

1K