News for hackers. Stuff that matters.

m3affan

But how good are these models compared to gpt4o? My last experience with llama2-8b was not great at all. Are there really that good models that would fit on an average consumer hardware (mine has already 32GB ram and 16GB vram)?
ChatGPT zealots are still hating on that piece.
Re:Faster Neural Networks Straight from JPEG on Thursday July 13, 2023 @02:21PM (#36713786)
Transformers are the taking the cake in the Deep Learning community
Re:InternLM – new open source 7B LLM on Thursday July 06, 2023 @07:05AM (#36614069)
Interesting attack vector. Malicious model codes.
Re:A curated list of uBlock origin filters on Tuesday July 04, 2023 @11:45AM (#36588211)
Awesome work. It's sad to see the internet being ridden by cancerous ads more and more.
But that's the reality. Doing good on an interview does not guarantee the ability to do the job
"But, but, it's illegal". But nonetheless it was a illuminating debate.
Re:Stanford A.I. Courses on Tuesday July 04, 2023 @02:36AM (#36583002)
Karpathy is great researcher in the field, but that doesn't always translate into good explanation skills automatically.
Re:Leaking Bitwarden's Vault with a Nginx vulnerability on Tuesday July 04, 2023 @12:10AM (#36582139)
What are the alternarives though?
I wonder how many concepts that are so complicated for our brain to formalize or even process.
Re: on Sunday July 02, 2023 @03:29AM (#36559154)
This sounds very US centric. In the EU things work differently.
Re:Teach yourself Computer Science functionally on Wednesday June 14, 2023 @02:41AM (#36322273)
Thanks for the references!
This was an amazing talk. He did a great job keeping it entertaining.
Re:Llama.cpp: Full CUDA GPU Acceleration on Tuesday June 13, 2023 @08:14AM (#36309098)
This is an impresdive use case
Re:Make your own VPN with Fly.io, tailscale and GitHub on Thursday May 25, 2023 @07:30AM (#36069382)
But, but ... Nordvpn
Re:Make your own VPN with Fly.io, tailscale and GitHub on Thursday May 25, 2023 @07:28AM (#36069364)
But, but ... nordvpn?
Re:QLoRA: Efficient Finetuning of Quantized LLMs on Thursday May 25, 2023 @07:27AM (#36069350)
Thanks for pointing this out. Such projects deserve visibility.
Exactly this. I am pretty sure he's bluffing, and even if not, by the time they decide to pull their operations, many alternative would be on the market (there are already many).
The hero we need
Re:Faster CPython at PyCon, part one on Thursday May 11, 2023 @04:47PM (#35908143)
This guy cythons
Databricks is on a roll
Re:LoRA: Low-Rank Adaptation of Large Language Models on Friday March 24, 2023 @05:39PM (#35295968)
Let the revolutionbbegin
Re:GPT-4 on Tuesday March 14, 2023 @05:47PM (#35159359)
The power openai will hold above everyone else is just too much. They will not allow their AI as a service without data collection. That will be a big pill to swallow for the EU.
Re:Bing: “I will not harm you unless you harm me first” on Thursday February 16, 2023 @08:48AM (#34818519)
It's gotten so bad that it's hard to even authenticate Bing prompts.
Re:The Odin – DIY genetic engineering on Monday February 13, 2023 @12:48PM (#34777198)
I am curious how such business model arose?
I wonder how lasting will the support be for such libraries
Re:DIY 1,500W solar power electric bike on Thursday February 09, 2023 @04:25AM (#34721550)
What were the costs?
Re:New AI classifier for indicating AI-written text on Tuesday January 31, 2023 @01:52PM (#34599876)
There is work on hidden signatures in generated text, invisible to humans. Only way to move forward.
Re:Build ChatGPT like chatbots on your website on Sunday January 29, 2023 @12:53PM (#34570650)
On that note, have there been any chatgpt like open-source projects, that are on the same or similar level?
Re:How a CPU works: Bare metal C on my RISC-V toy CPU on Thursday January 26, 2023 @03:58PM (#34538736)
Mvp, thanks for sharing

Older »