Hacker News: sieszpak

New comment by sieszpak in "I stopped using AI code editors"

sieszpak — Thu, 03 Apr 2025 06:37:36 +0000

I’m using Sonnet 3.7 now, previously used Sonnet 3.5, and I recommend it. If you know what you want to do, the work becomes pleasant and orders of magnitude faster. IMO, you need to learn how to code from scratch with LLMs. Remember—garbage in, garbage out. The Pareto principle also applies here: AI does 80% in the first shot.

UsAsk HN: What should I use to build UI interface for an LLM

sieszpak — Thu, 25 Jul 2024 20:29:02 +0000

I want to create an LLM UI interface for my fine-tuned Llama model, similar to a private GPT chat. The interface should allow me to upload multiple documents, add prompts to them, capture the response, and format it with Markdown tags based on one of the attached template documents. What should I use? Preferably something open-source so I can customize it. What do you recommend? I looked at Gradio, but it seems too “modest,” and OpenWebUI is too complex and extensive. I would also like the solution to support PWA technology.

Comments URL: https://news.ycombinator.com/item?id=41073026

Points: 1

# Comments: 1

New comment by sieszpak in "Jacek Karpińśki, the computer genius the communists couldn't stand (2017)"

sieszpak — Thu, 25 Jul 2024 20:01:24 +0000

Many Poles have distinguished themselves in various fields of science and life. I recognize Madame Curie, Copernicus, Ulam, Pope John Paul II, Chopin, Polanski, and perhaps Wozniak. I had a Polish professor at Mellon University who talked a bit about it, and there was a lot.

New comment by sieszpak in "Cognitive psychology tests show AIs are irrational"

sieszpak — Wed, 05 Jun 2024 09:36:00 +0000

You're right. However, irrational LLMs will provide better medical and psychological advice than human experts and demonstrate much higher levels of empathy. How to combine it now - irrationally and it will replace some of the psychologist...

New comment by sieszpak in "Cognitive psychology tests show AIs are irrational"

sieszpak — Wed, 05 Jun 2024 09:15:55 +0000

Unfortunately, researchers do not understand how LLMs work and they attribute their common known defects resulting from, among others, tokenization processes as irrational actions from a psychological point of view. It is known that the models cannot cope with mathematical operations or division into consonants and vowels - and these are the most serious allegations in the article... a weak article, written by dilettantes.

New comment by sieszpak in "Duolingo removes all LGBT content from its app in Russia"

sieszpak — Wed, 05 Jun 2024 09:08:07 +0000

Pecunia non olet

New comment by sieszpak in "FineWeb: Decanting the web for the finest text data at scale"

sieszpak — Mon, 03 Jun 2024 20:09:19 +0000

https://www.youtube.com/watch?v=20wbA_ijjmg&t=696s

New comment by sieszpak in "FineWeb: Decanting the web for the finest text data at scale"

sieszpak — Mon, 03 Jun 2024 10:58:54 +0000

The article contains a recipe worth millions of dollars. And there's silence in the comments... why?

New comment by sieszpak in "FineWeb: Decanting the web for the finest text data at scale"

sieszpak — Sun, 02 Jun 2024 13:19:11 +0000

Any ideas for a "magic sauce" for this?

New comment by sieszpak in "FineWeb: Decanting the web for the finest text data at scale"

sieszpak — Sun, 02 Jun 2024 13:17:47 +0000

I agree, the information provided in this article is a treasure. Maybe someone will add some "magic sauce" to it?

New comment by sieszpak in "What We Learned from a Year of Building with LLMs"

sieszpak — Wed, 29 May 2024 06:08:33 +0000

I would like to know your opinion about grafRAG and the ontology. Knowledge Graphs (KG) are a game changer for companies with a lot of unstructured data in the context of applying them with LLM

New comment by sieszpak in "LLaMA 3 70B Llamafiles"

sieszpak — Sat, 20 Apr 2024 14:53:38 +0000

Llama 3 8B seems sad to answer... this is the first model in a long time that has had trouble telling me how much is 3! - (factorial)

New comment by sieszpak in "Intermittent fasting linked to a 91% higher risk of cardiovascular death"

sieszpak — Tue, 19 Mar 2024 09:55:55 +0000

Are pharmaceutical companies starting to see a decline in sales? Recently, the process of autophagy was praised, and its discoverer even received a Nobel Prize. Starting today, I'm beginning fasting because if they want to convince me that it kills, I'm now certain that this is another lie.

New comment by sieszpak in "Intermittent fasting tied to higher risk of cardiovascular death, research shows"

sieszpak — Tue, 19 Mar 2024 09:55:31 +0000

New comment by sieszpak in "Fine tune a 70B language model at home"

sieszpak — Fri, 08 Mar 2024 14:02:44 +0000

4x 3080???