Sign in

Duarte

@duarteocarmo.com
987 followers 232 following 425 posts

Born in 🇵🇹, based in 🇩🇰. Machine Learning Engineer (or whatever it's called now) duarteocarmo.com duarteocarmo.com/photos github.com/duarteocarmo/

PostsRepliesMedia
Duarte @duarteocarmo.com · 5h
Wonderful two days for Denmark. CPR? Hacked. DTU? Hacked as well? Is this a different attack? JESUS.
000
Duarte @duarteocarmo.com · 01/10/2026
LLM as a judge is a great idea!
210
Duarte @duarteocarmo.com · 24/09/2026
I hate you 😂
130
Duarte @duarteocarmo.com · 22/09/2026
That's what i thought
000
Duarte @duarteocarmo.com · 18/09/2026
Increasingly telling models that do "long horizon tasks"
000
Duarte @duarteocarmo.com · 17/09/2026
Invoice dataset is so useful that all the invoices look the same 😂
000
Duarte @duarteocarmo.com · 15/09/2026
I saw a post on LinkedIn saying we should use AMALIA to grade Portuguese high-school exams (the gov. was taking too long) Horrible idea.. But! I benchmarked a bunch of smaller models vs. AMALIA on a bunch of EU-PT benchmarks (including PT-exams) link below
100
Duarte @duarteocarmo.com · 30/08/2026
"Ping!" Agent ready. Running 2-3 agents in parallel doesn't need to be mentally taxing. Pi-jumper allows you to jump between @pidotdev sessions running on tmux with a couple clicks. It's part of pi-tools, some extensions I've built around Pi! Blog post below
100
Duarte @duarteocarmo.com · 25/08/2026
Presenting Bagaço v3 the next generation of Bagaço, the largest pre-training dataset for European Portuguese! ~30 Billion tokens from across the Web, it now also includes Wikipedia and PDF data thanks to the great work from the folks at Hugging Face! Full write up in the comments
110
Duarte @duarteocarmo.com · 19/08/2026
Build a herdr competitor for @pidotdev in 30 seconds today. @ycombinator please invest
100
Duarte @duarteocarmo.com · 07/08/2026
This year I've built evals, classifiers, and even Bagaço, a pretraining dataset for European Portuguese. This raises an obvious question: Can we pretrain LLMs fully in European Portuguese? How do we measure if they're any good? Meet project Ginjinha 🇵🇹 (link in comments)
230
Duarte @duarteocarmo.com · 04/05/2026
This month in Retrospectiva: Using Hermes as a running coach and travel agent, Carlo Rovelli, Jensen Huang, and the Plastic Detox. duarteocarmo.com/blog/retros...
010
Duarte @duarteocarmo.com · 05/01/2026
I don't like ads. You don't like ads. We don't like ads. Limpa is like AdBlock + SponsorBlock but for podcasts. GitHub repo below open.substack.com/pub/duarteo...
120
Duarte @duarteocarmo.com · 02/01/2026
Now with agents you can have super meaningful commit messages. It's incredible. Lul
000
Duarte @duarteocarmo.com · 01/01/2026
Dirty little secret. Cancel Claude code, cancel ChatGPT plus. Github copilot is 10 bucks a month. And OpenCode can use them all 🤫
000
Duarte @duarteocarmo.com · 31/12/2025
2000 km for 2025 ✅
000
Duarte @duarteocarmo.com · 30/12/2025
Github actions in a nutshell: Comment in an issue -> get a notification for a failed action 😂
000
Duarte @duarteocarmo.com · 19/12/2025
Do I have to come here everytime a new model comes out?
030
Duarte @duarteocarmo.com · 18/12/2025
When do they send the badge? @OpenRouterAI
010
Duarte @duarteocarmo.com · 16/12/2025
"Mathematician extraordinaire" Or, I probably shouldn't use @raycast to do basic math so much 😂
020
Duarte @duarteocarmo.com · 16/12/2025
Denmark has a map of all the different AI projects being developed in the different municipalities It's crazy to me sometimes how fast this country moves 😂 www.kl.dk/videncenter...
010
Duarte @duarteocarmo.com · 16/12/2025
Should rethink my branch names?
000
Duarte @duarteocarmo.com · 11/12/2025
Can relate
000
Duarte @duarteocarmo.com · 10/03/2025
Late last year I got the opportunity to present at the very first PyCon Wroclaw. I talked about solving the right problem, the state of text-to-SQL, and systematically improving your question-answering system with evaluations. Oh, and memes. We gotta have those duarteocarmo.com/blog/you-als...
010
Duarte @duarteocarmo.com · 09/01/2025
Danish winter cheatcodes
000
Duarte @duarteocarmo.com · 02/01/2025
Here's an interesting problem to test different LLMs at solving. Take this image, and try to create a python function that solves it. I tried deepseek v3, o1, most models. The one that got closest was 3.5 sonnet. But there a bug. Care to find it?
000
Duarte @duarteocarmo.com · 29/12/2024
Lisbon's most annoying problem. 8 hours by train won't even get you to Madrid.
010
Duarte @duarteocarmo.com · 29/12/2024
Sunday news 😂
000
Duarte @duarteocarmo.com · 28/12/2024
Reason I like coming to Italy #24
120
Duarte @duarteocarmo.com · 26/12/2024
My favorite gadget purchase in a long while. Very tempted to write a review! (Will likely be more of a homage)
Boox palma
360
Duarte @duarteocarmo.com · 25/12/2024
Christmas hoodie upgrade.
020
Duarte @duarteocarmo.com · 19/12/2024
Best memes. (stolen from @mlopscommunity.bsky.social)
131
Duarte @duarteocarmo.com · 18/12/2024
The message you dont want to receive when you leave your place for the holidays.
010
Duarte @duarteocarmo.com · 18/12/2024
They keep coming. I'm still absolutely clueless.
010
Duarte @duarteocarmo.com · 17/12/2024
Denmark weather decides to give me sun the day I fly south. Classic.
130
Duarte @duarteocarmo.com · 17/12/2024
Mistery of the day. 34 signups for infrequent.app in the last 12 hours. My analytics show 9 unique visitors in the last 7 days. What's going on here? (Already checked and plausible appears to be working fine..)
220
Duarte @duarteocarmo.com · 17/12/2024
Fun
000
Duarte @duarteocarmo.com · 13/12/2024
Morning wisdom from @chiphuyen.bsky.social
040
Duarte @duarteocarmo.com · 10/12/2024
Here we go again.
1 Termal Bologna Marathon
Duarte Carmo l
, Step 3
• Termal Bologna Marathon | Bologna Marathon - Atleti stranieri non
tesserati (non competitiva)
Duarte Carmo
• Commissioni di servizio
Duarte Carmo, 0.90 € + 6.00 %
010
Duarte @duarteocarmo.com · 10/12/2024
Becoming a pattern..
040
Duarte @duarteocarmo.com · 07/12/2024
Laptop got taken away from me
Beers bar
010
Duarte @duarteocarmo.com · 07/12/2024
Lol you sound like me in my Wroclaw talk last Saturday kkk duarteocarmo.com/pdfs/pycon-w...
110
Duarte @duarteocarmo.com · 07/12/2024
None of this needed. You can really raw dog it, but I like the one from thoughtworks: www.thoughtworks.com/insights/blo...


    Use benchmarks (with grains of salt) to compare model capabilities directly

    Use evals to measure and understand the performance characteristics of your system

    Use tests to validate and action upon these learnings (e.g. “fail” if some set of metrics dip below acceptable thresholds)
000
Duarte @duarteocarmo.com · 04/12/2024
I like this Nova model! Worth a try imo.. Attractive pricing and pretty fast.
121
Duarte @duarteocarmo.com · 02/12/2024
This is going to go so badly.
020
Duarte @duarteocarmo.com · 02/12/2024
Another day - another approved model!
000
Duarte @duarteocarmo.com · 01/12/2024
This is going to be a good one. Filled with pain aha
000
Duarte @duarteocarmo.com · 01/12/2024
This is how git is done
150
Duarte @duarteocarmo.com · 30/11/2024
Denmark 0, Poland 1 Find the odd one out
010
Duarte @duarteocarmo.com · 29/11/2024
Vibes check passed
000