Tweet épinglé
Deep
15.8K posts
Deep
@DeepStarts
20. Developer. @ Intern x3. l Helping brands in product launch | Dm for collab/work
DM for collab📥 Inscrit le Ağustos 2025
2.2K Abonnements5.4K Abonnés
🚨 Open-source AI keeps getting better.
Ornith-1.0 just dropped:
-> 397B MoE
-> 82.4 SWE-Bench Verified
-> 77.5 Terminal-Bench
-> MIT License
The most interesting part?
It doesn’t just learn solutions.
It learns how to create better coding workflows for itself.
Ornith@ornith_
Aloha! 🌺 Meet Ornith-1.0, a family of open-source LLMs specialized for agentic coding. Ornith-1.0 spans the full parameter sizes including 9B Dense, 31B Dense, 35B MoE, and 397B MoE. It achieves state-of-the-art performance among open-source models of comparable size on coding benchmarks including: ✅Terminal-Bench 2.1(77.5) ✅SWE-Bench(82.4 on verified, 62.2 on pro, 78.9 on Multilingual) ✅NL2Repo(48.2) ✅SWE Atlas(41.2 on QnA, 42.6 RF, 39.1 TW) ✅ClawEval(77.1) Post-trained on top of gemma4 and qwen3.5, Ornith-1.0 employs a novel self-improving training strategy in which reinforcement learning is used to generate not only solution rollouts, but also the task-specific scaffolds that drive those rollouts. By jointly optimizing the scaffold and the resulting solution, the model generate higher-quality solutions in agentic coding.😎 All models are released under the MIT license, enabling full commercial and research use. 📖Tech Blog: deep-reinforce.com/ornith_1_0.html 🤗Huggingface: huggingface.co/collections/de…
English
Most tech billionaires weren't born in boardrooms.
They started by writing code.
Bill Gates built software.
Mark Zuckerberg coded Facebook in a dorm.
Larry Page & Sergey Brin built Google's search engine.
Jeff Bezos wrote software on Wall Street before Amazon.
Elon Musk coded Zip2 and X .com.
Michael Dell built PCs and software systems from his dorm.
Turns out the fastest path to building a billion-dollar tech company often starts with a blinking cursor and a lot of caffeine. 💻
English
My 2 friends started their careers at the same time.
Rohan moved to Bangalore-
> Salary: ₹17 LPA
> Rent: ₹35,000/month
> Daily commute
> Expensive food & weekends
> Savings after a year: ~₹1–2 lakh
Aman stayed in Indore-
> Salary: ₹8 LPA
> Rent: ₹6,000/month
> Home-cooked food
> Zero commute
> Savings after a year: ~₹4–6 lakh
Who's actually winning and why?
English
Unpopular opinion: The MacBook Touch Bar deserved improvements and new features, not complete removal.
English
Name a profession that you believe is most future-proof against AI ?
English
Top 20 Tech Roles and what they actually do:
• 🧠 AI Engineer — Copy-pasting prompts until the chatbot sounds polite.
• 🤖 MLOps Engineer — Keeping the AI alive on a mountain of expensive GPUs.
• 🏗️ DevOps Engineer — Turning it off and on again, but at a massive scale.
• ⚡ Site Reliability (SRE) — Staring at red dashboards at 3 AM while drinking cold coffee.
• 🎯 Product Manager — Scheduling meetings to discuss when to schedule the next meeting.
• 🛡️ Cybersecurity Analyst — Begging coworkers to stop clicking on Free iPad emails.
• 💻 Full-Stack Engineer — Pretending to understand both the backend and frontend equally.
• 🧱 Backend Engineer — Writing beautiful code that breaks because the database sneezed.
• 🎨 Frontend Developer — Centering a div element. (Still trying).
• 📊 Data Scientist — Making colorful charts to justify decisions executives already made.
• 🗄️ Data Engineer — Plumbing pipeline pipes that overflow with bad spreadsheet data.
• 🔍 QA Tester — Purposefully breaking things and watching developers cry.
• 👔 Tech CTO — Saying the word AI 45 times in a 10-minute presentation.
• 🤝 Scrum Master — Professional digital sticky note mover.
• 🎨 UI/UX Designer — Moving a button 2 pixels to the left, then moving it back.
• 📱 Mobile Developer — Constantly fighting the Apple App Store review team.
• ☁️ Cloud Architect — Explaining why the server bill is suddenly higher than rent.
• 🔧 Systems Administrator — The only person who knows why the office printer actually works.
• 📦 Product Owner — Adding 500 minor features to the backlog that will never get built.
• 📉 Data Analyst — Aggregating data into a pivot table that someone will format into a PDF.
Which one did I target too accurately?
English
What's even the point of jumping to Linux if Chrome is still your first install?
English
