Nick Burns
@nickdaleburns
Followers
564
Following
36K
Media
138
Statuses
3K
Data scientist, avid squash player, coffee drinker and ML Engineer. Working on self-supervised methods for vision and healthcare.
Joined March 2018
PASSION - A strong liking, desire, or devotion to an activity. When I started playing tennis at 8 years old, it was just a game. Then it became my passion. My dream was to one day became a professional tennis player. I know that as an athlete, people like to think they know
554
970
11K
@NickSmith_MP @rachelboyack @NelTasChamberNZ this is the type of thing that regional councils and representatives should be making noise about.
0
0
0
Could you imagine the negative impact to Motueka and @KaiteriteriNZ @KaiteriKayak @AbelTasmanNP started charging for parking at Kaiteriteri or Mārahau? It doesn't stack up.
0
0
0
$1 million is absolute chump change to the government. Guaranteed that their forecasts don't include lower visitor numbers as a result, or reduced tourism satisfaction and flow on impacts. https://t.co/i57zU1LHJM
rnz.co.nz
The first 20 minutes will be free with a rate of $5 per hour after.
2
0
0
upgrade your transformers 🔥 it comes with insanely capable models like SAM2, KOSMOS2.5, Florence-2 and more 🤝 I built a notebook you can run with free Colab T4 to walk through the API for new models 🙋🏻♀️ fine-tuning will follow-up soon!
14
24
235
Ever wanted to jump into a satellite image? Here are some Nano Banana perspectives on maps 🧵
4
1
8
Really excellent work here! Random forest and XGBoost are overused, overhyped, and likely lead to worse decision making than if not used. That these are to facilitate “precision medicine” is a misunderstanding of the techniques and curse of dimensionality @RWJE_BA @f2harrell
In our latest work, we demonstrate that risk estimates for patients are HUGELY uncertain due to model uncertainty, data uncertainty, and population uncertainty. Even when based on large sample sizes. @laure_wynants @ESteyerberg
https://t.co/clSxKPyu1m
3
13
54
we just released another chapter on O'Reilly for our vision language model book about modern VLM architectures ✨ @andimarafioti @micuelll @orr_zohar it covers early & late fusion, encoder-decoders, multimodal attention types and more! 🤗
6
20
242
Zero shot classification with DinoV3
3
1
11
Stolen from Silver Snuffull Wolf on FB What I’d say to you is I’m putting it to you that what I’m saying is I don’t know
16
114
291
This is a desperate, exploitative move from the government. Won't achieve anything except discouraging tourism. And how on earth do they enforce this without it costing hundreds per visitor? More users pay from a government out of ideas.
What next! Toll booths on road to Mt Cook & at entrance to Tongariro Crossing for international tourists who already pay a lot to get to & travel in NZ & also pay a visitor levy already? More commercial exploitation of conservation land? Desperate stuff:
0
0
2
I'm no ACT voter. But Green feels equally driven by identity politics. I'm probably a typical centrist NZ voter. Voting for stability. Hopefully incremental long-term positive steps. I'm deeply worried by the unchecked swing we're currently seeing. Great article from Vance.
A sad commentary on state of #NZ politics today: “The political pendulum doesn’t swing. Instead, wrecking ball politics now smashes through the written & unwritten codes that once held together democracy”. Time to rediscover that we have more in common….
0
0
2
A sad commentary on state of #NZ politics today: “The political pendulum doesn’t swing. Instead, wrecking ball politics now smashes through the written & unwritten codes that once held together democracy”. Time to rediscover that we have more in common….
thepost.co.nz
OPINION: Compromise is out, consensus is dead, and the country is being governed by tantrum and teardown.
5
93
368
TabM now has a Python package! TabM is a simple and powerful DL architecture for tabular data that efficiently imitates an ensemble of MLPs 🏆 TabM has been used in winning solutions on Kaggle, and performs well on TabReD -- a challenging benchmark! 💻 pip install tabm 👇Link
3
54
284
You can just continue pre-train things ✨ Happy to announce the release of BioClinical ModernBERT, a ModernBERT model whose pre-training has been continued on medical data The result: SOTA performance on various medical tasks with long context support and ModernBERT efficiency
Very excited to share the release of BioClinical ModernBERT! Highlights: - biggest and most diverse biomedical and clinical dataset for an encoder - 8192 context - fastest throughput with a variety of inputs - sota results across several tasks - base and large sizes (1/8)
4
33
214
Our economy is insufficiently productive because we are structurally and socially encouraged to make money by buying and selling houses to each other. Successive govts have been out of ideas, and short on courage.
nzherald.co.nz
OPINION: Gross debt is projected to increase by $73b, reaching $283b by 2029.
0
1
6
Why do they call it a model? Because it models something. What does modeling something mean? Creating a simplified representation of something. Whose simplified representation is it? Data's. If you understand this simple logic, you understand everything about LLMs: what they
20
22
201