From e6dcfb12c15cf407cc2b1fb7ae59457b01b410eb Mon Sep 17 00:00:00 2001 From: Paul Buetow Date: Sun, 31 May 2026 22:32:36 +0300 Subject: Update content for html --- gemfeed/atom.xml | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'gemfeed/atom.xml') diff --git a/gemfeed/atom.xml b/gemfeed/atom.xml index 2606a48d..e6ae257e 100644 --- a/gemfeed/atom.xml +++ b/gemfeed/atom.xml @@ -1,6 +1,6 @@ - 2026-05-31T18:38:04+03:00 + 2026-05-31T22:32:28+03:00 foo.zone feed To be in the .zone! @@ -15,14 +15,14 @@ Paul Buetow aka snonux paul@dev.buetow.org - I created a calculator. Not because the world needed another one, but because I wanted to test something: how well do local LLMs hold up as pair programmers on a real project? + I created a calculator. Not because the world needed another one, but because I wanted to test something: How well do local LLMs hold up as pair programmers on a real project? I want to be independent of the big LLM providers like Anthropic, OpenAI, and company for this `gt`-project.

gt calculator - a calculator built with local LLMs



Published at 2026-05-31T14:24:10+03:00

-I created a calculator. Not because the world needed another one, but because I wanted to test something: how well do local LLMs hold up as pair programmers on a real project?
+I created a calculator. Not because the world needed another one, but because I wanted to test something: How well do local LLMs hold up as pair programmers on a real project? I want to be independent of the big LLM providers like Anthropic, OpenAI, and company for this gt-project.

The answer is: well enough for small projects like this.

@@ -33,7 +33,7 @@ gt on Codeberg
gt logo

-The whole thing — code, tests, documentation, even the logo — was built using only LLMs that can run locally on reasonable hardware: Qwen, Gemma, Nemotron, GPT-OSS. To be honest, I didn't run them locally either — I rented a Hyperstack VMs with NVidia GPU just to get a feel for the quality before investing in hardware. The point was to test models that don't require a cloud API and could realistically run on your own box.
+The whole thing — code, tests, documentation, even the logo — was built using only LLMs that can run locally on reasonable hardware: Qwen, Gemma, Nemotron, GPT-OSS. To be honest, I didn't run them locally either — I rented Hyperstack VMs with NVidia GPU just to get a feel for the quality before investing in hardware. The point was to test models that don't require a cloud API and could realistically run on your own box.

https://www.hyperstack.cloud/

-- cgit v1.2.3