Topic
21 articles found
Half a year ago, I wanted to see for myself what can we currently have with local LLMs. I went down the rabbit hole, learned quite a lot in the process, and shared my results in an article. The results were pretty discouraging: even with 32 GB VRAM, the best models I could run were both too slow and
You tell your AI “Polish my email and send it.” Same sentence, three outcomes. The gap between Large Action Models (LAMs) and agentic LLMs is one of the most practically important distinctions in AI today, and also one of the least clearly explained. In this article, we cut through the confusion thr
Most "LLM wikis" use agents, embeddings, and repeated model calls to organize local notes. I built a deterministic alternative: a pure Python compiler that turns messy markdown into a linked, linted wiki using only the standard library. Along the way, I fixed two real bugs, benchmarked the pipeline
t0-alpha is a decoder-style patch transformer for probabilistic time-series forecasting. Raw series are split into 32-step patches, embedded, processed through causal time-attention and group-attention layers, and decoded into future quantiles rather than a single point forecast. The post Time-Serie
This article explains why traditional CI/CD gates are not enough for production AI systems. I share a practical release-gating approach The post Why traditional CI/CD fails for LLMs (and the release gates we built to fix it) appeared first on The New Stack.
Let’s start with a game. Open up your chatbot of choice—Claude, ChatGPT, Gemini—and type “Give me a random number between 1 and 10.” You’re going to get 7. Almost always. Now type “Another” and you’ll get 3 or 4. Type “Another” again and you’ll get 8 or 9. That won’t work every time—but if it…
My smart home camera alerts used to be useless, but now they tell me what actually happened
Hey everyone, I wanted to share a quick reflection on how much the field of AI research has changed recently and see how others are navigating it. Looking at the landscape today, there has been a massive shift since LLMs and VLMs took over the market. It used to be that the work was centered around
After months of testing local LLMs, I found that productivity depends on tools, not just models.
Beat the 8GB VRAM limit. Learn how to run three different LLMs on a single 8GB GPU using C++ layer multiplexing and admission control. The post 3 Agents. 3 LLMs. 1 Aging GPU: Engineering Parallel Inference on Bare Metal appeared first on Towards Data Science.
Enterprise Document Intelligence [Vol.1 #7C] - One LLM call ranks the candidates with reasons. The output is one typed object your auditor can defend The post Letting an LLM Pick the Right RAG Page: The Arbiter Pattern at the End of Retrieval appeared first on Towards Data Science.
Google's GKE Labs has introduced OpenRL, an open-source project that provides a self-hosted API for post-training and fine-tuning Large Language Models (LLMs) on standard Kubernetes clusters. By Sergio De Simone
Has anyone able to run the code from serving-llms-efficiently-with-vllm-–-part-ii in kaggle or colab. I am unable to complete the steps mentioned in here 1 post - 1 participant Read full topic
How much of Nvidia is reflected by their visionary CEO, Jenson Huang? With his praise and later support of OpenClaw, The post “An agent is an LLM and a harness”: What Nvidia really thinks about OpenClaw appeared first on The New Stack.
TL;DR Over 844,000 sites had implemented llms.txt by late 2025 — still under 1% of active websites, meaning early adoption is a differentiator; Cloudflare, Anthropic, Stripe, and Vercel all ship one. HTML is a lossy agent interface: every step from fetch to context extraction degrades content; llms.
This is a cross-post. The original — with the full module-by-module comparison table, feature breakdown, and FAQ — is on angeo.dev. AI systems are increasingly shaping how users discover products. Instead of browsing category pages, users now ask: "best hiking boots under €150" and get a synthesized
I built a small web scraping framework in Rust, mostly with an AI doing the typing. It's called ferrous — a Colly-style collector: register CSS selector callbacks, queue URLs, write JSONL. About 700 lines. The pitch I kept hearing, and half-believed, was that Rust and LLMs are a good match now: the
It ain't no match for a dedicated GPU, but you can run some light LLMs on the N100
Stop Pasting Your Code Into ChatGPT For Debugging—Run LLMs Locally Instead Here's the scenario: You've got a nasty bug, and your first instinct is to copy the suspicious function into ChatGPT. Works great. Except now you've just sent your company's code, your API keys (if you weren't careful), and p
It's the holy trinity of cost savings when it comes to LLMs
