unclecode's profile picture. Author of Crawl4AI (#1 GitHub Trending). Founder of Kidocode, SE Asia's largest tech & biz school. Leading AI research on synthetic data, AI consultant.

Unclecode (Hossein)

@unclecode

Author of Crawl4AI (#1 GitHub Trending). Founder of Kidocode, SE Asia's largest tech & biz school. Leading AI research on synthetic data, AI consultant.

고정된 트윗

Crawl4ai crossed 50K start 🌟 (~5k forks). Thanks to all who shared their love and support. A project I started to build a tool for my company turned into a lovable open-source project at this scale. It's so rewarding to be part of an authentic change. I like to believe Crawl4ai…

unclecode's tweet image. Crawl4ai crossed 50K start 🌟 (~5k forks). Thanks to all who shared their love and support. A project I started to build a tool for my company turned into a lovable open-source project at this scale. It's so rewarding to be part of an authentic change. I like to believe Crawl4ai…

Opus 4.1 now is "Legacy"?!

unclecode's tweet image. Opus 4.1 now is "Legacy"?!

Unclecode (Hossein) 님이 재게시함

New paper: You can make ChatGPT 2x as creative with one sentence. Ever notice how LLMs all sound the same? They know 100+ jokes but only ever tell one. Every blog intro: "In today's digital landscape..." We figured out why – and how to unlock the rest 🔓 Copy-paste prompt: 🧵


Opus to design, Sonnet to implement, Haiku to maintain

Introducing Claude Haiku 4.5: our latest small model. Five months ago, Claude Sonnet 4 was state-of-the-art. Today, Haiku 4.5 matches its coding performance at one-third the cost and more than twice the speed.

claudeai's tweet image. Introducing Claude Haiku 4.5: our latest small model.

Five months ago, Claude Sonnet 4 was state-of-the-art. Today, Haiku 4.5 matches its coding performance at one-third the cost and more than twice the speed.


NVIDIA DGX Spark finally is available. It has ~100 TFLOPs of FP16 with 128GB of memory at 273 GB/s bandwidth. The Mac Studio has 512GB of unified memory at 819 GB/s, but the GPU only has ~26 TFLOPs of FP16. So one is faster in compute and one better in memory bandwidth. The DGX…

Clustering NVIDIA DGX Spark + M3 Ultra Mac Studio for 4x faster LLM inference. DGX Spark: 128GB @ 273GB/s, 100 TFLOPS (fp16), $3,999 M3 Ultra: 256GB @ 819GB/s, 26 TFLOPS (fp16), $5,599 The DGX Spark has 3x less memory bandwidth than the M3 Ultra but 4x more FLOPS. By running…

exolabs's tweet image. Clustering NVIDIA DGX Spark + M3 Ultra Mac Studio for 4x faster LLM inference.

DGX Spark: 128GB @ 273GB/s, 100 TFLOPS (fp16), $3,999
M3 Ultra: 256GB @ 819GB/s, 26 TFLOPS (fp16), $5,599

The DGX Spark has 3x less memory bandwidth than the M3 Ultra but 4x more FLOPS.

By running…


Finally, GIL is not the boss! Can’t imagine how this will ripple through everything built on Python, AI, web, data, even infra. The entire stack just leveled up!

Finally, Python 3.14 lets you disable GIL! It's a big deal because earlier, even if you wrote multi-threaded code, Python could only run one thread at a time, giving no performance benefit. But now, Python can run your multi-threaded code in parallel. And uv fully supports it!



MIT License

This is insane 🤯 A new system called Paper2Video can read a scientific paper and automatically create a full presentation video slides, narration, subtitles, even a talking head of the author. It’s called PaperTalker, and it beat human-made videos in comprehension tests.…

aiwithmayank's tweet image. This is insane 🤯

A new system called Paper2Video can read a scientific paper and automatically create a full presentation video  slides, narration, subtitles, even a talking head of the author.

It’s called PaperTalker, and it beat human-made videos in comprehension tests.…


Before AI human publishing “hallucination articles and posy” almost every day or even hour in social media and web, just we couldn’t make em pay $440k! Now for AI “hallucination” every one is hyping to stop them. From one side, we keep trying to mimic human intelligence, then…

So surprising. They should have at-least proof-read. 😅 AI hallucinations need to be SOLVED ASAP. 💰 Deloitte has agreed to repay part of a $440,000 government fee after admitting it used AI to write sections of an official report that later turned out to have fake references…

rohanpaul_ai's tweet image. So surprising. They should have at-least proof-read. 😅

AI hallucinations need to be SOLVED ASAP. 

💰 Deloitte has agreed to repay part of a $440,000 government fee after admitting it used AI to write sections of an official report that later turned out to have fake references…
rohanpaul_ai's tweet image. So surprising. They should have at-least proof-read. 😅

AI hallucinations need to be SOLVED ASAP. 

💰 Deloitte has agreed to repay part of a $440,000 government fee after admitting it used AI to write sections of an official report that later turned out to have fake references…


I remember Sergey Brin, was saying all models do better if they are threatened, and this research shows 4.8% more average accuracy with more rude prompts! 😳

unclecode's tweet image. I remember Sergey Brin, was saying all models do better if they are threatened, and this research shows 4.8% more average accuracy with more rude prompts! 😳

Rude prompts to LLMs consistently lead to better results than polite ones 🤯 The authors found that very polite and polite tones reduced accuracy, while neutral, rude, and very rude tones improved it. Statistical tests confirmed that the differences were significant, not…

rohanpaul_ai's tweet image. Rude prompts to LLMs consistently lead to better results than polite ones 🤯 

The authors found that very polite and polite tones reduced accuracy, while neutral, rude, and very rude tones improved it.

Statistical tests confirmed that the differences were significant, not…


Crawl4ai Version 0.7.5 is coming 😎 my fav feature, pass hooks to the docker api, you can literally pass python function to the api 🔥🔥🔥 this was a request since the very beginning of this journey, sorry took time, but its coming soon! #crawl4ai

unclecode's tweet image. Crawl4ai Version 0.7.5 is coming 😎 my fav feature, pass hooks to the docker api, you can literally pass python function to the api 🔥🔥🔥 this was a request since the very beginning of this journey, sorry took time, but its coming soon! 
#crawl4ai

Unclecode (Hossein) 님이 재게시함

got a cheap e-ink display, coded up a dashboard, and installed a weather station on the roof. first rain today and seeing the rainfall totals update in real time is pure nerd dopamine

kyleturman's tweet image. got a cheap e-ink display, coded up a dashboard, and installed a weather station on the roof. first rain today and seeing the rainfall totals update in real time is pure nerd dopamine

Very unfair! I started claude code before max plan, spending 1000$+ monthly, then switch to 200$, purchased team package for my team, using every day and never get “approaching to weekly limit” and now just in a short time see this! You guys never reply any message from people…

unclecode's tweet image. Very unfair! I started claude code before max plan, spending 1000$+ monthly, then switch to 200$, purchased team package for my team, using every day and never get “approaching to weekly limit” and now just in a short time see this! You guys never reply any message from people…

This is real Claude Code hits usage limits for the week in hours. I've never once hit their 5hr limit on the 20x plan I think this is the most egregious usage limit change thus far in the industry



This is really something, it make every laptop to become a freaking powerful Colab

Introducing Tinker: a flexible API for fine-tuning language models. Write training loops in Python on your laptop; we'll run them on distributed GPUs. Private beta starts today. We can't wait to see what researchers and developers build with cutting-edge open models!…

thinkymachines's tweet image. Introducing Tinker: a flexible API for fine-tuning language models.

Write training loops in Python on your laptop; we'll run them on distributed GPUs.

Private beta starts today. We can't wait to see what researchers and developers build with cutting-edge open models!…


I used Claude chrome extension, and as my first try, I gave it a reply I wanted to post for a video from @fireship_dev about n8n. It nicely done it, It gives me the same Claude Code vibe but in the browser, very impressed, good job @claudeai btw this is link to vide…

unclecode's tweet image. I used Claude chrome extension, and as my first try, I gave it a reply I wanted to post for a video from @fireship_dev about n8n. It nicely done it, It gives me the same Claude Code vibe but in the browser, very impressed, good job @claudeai btw this is link to vide…

Just added a "Copy Page" button to @crawl4ai docs! Now you can instantly copy any documentation page as markdown and feed it to your favorite AI coding assistant. Check it out → docs.crawl4ai.com #crawl4ai

unclecode's tweet image. Just added a "Copy Page" button to @crawl4ai docs! 
Now you can instantly copy any documentation page as markdown and feed it to your favorite AI coding assistant. 

Check it out → docs.crawl4ai.com

#crawl4ai

Unclecode (Hossein) 님이 재게시함

~2012 must have been like a crazy time. Neural networks were considered nonsense by most people. Hinton, LeCun and Bengio were the amongst the very few people who kept it all alive for decades. Alex Krizhevsky, a mad coding genius and a socially aloof kid, shows up to Hinton's…

madiator's tweet image. ~2012 must have been like a crazy time.

Neural networks were considered nonsense by most people. Hinton, LeCun and Bengio were the amongst the very few people who kept it all alive for decades.

Alex Krizhevsky, a mad coding genius and a socially aloof kid, shows up to Hinton's…

Unclecode (Hossein) 님이 재게시함

Excited to team up with @crawl4ai to make large-scale crawling and automation smoother. We are the powering captcha solving behind the scenes, making it more accessible and helping the community push boundaries forward🫂

Modern crawling often needs captcha solving at scale. @capsolver is now supporting and providing the backbone for captcha solving. They are creating examples, docs, and guides to make it easy for the community to use, and we will share them as well.



Loading...

Something went wrong.


Something went wrong.