Viewing profile — stefanwebb

I've created an alternative to CLI and MCP for locally running agentic tools. It uses Unix-based OS's named pipes, which means the client has quick IPC with the tool and it can hav…

First couple of paragraphs: "There are many things one needs to live a rich and fulfilled life (according to AI researchers). A good initialization [Mishkin and Matas, 2015], atten…

Here's a blog post I wrote last week on the same topic: https://blog.oumi.ai/p/small-fine-tuned-models-are-all-you I discuss a large-scale empirical study of fine-tuning 7B models …

Seems topical given some recent front-page HN articles on fine-tuning. I discuss a large-scale empirical study from 2014 of fine-tuning 7B models to outperform GPT-4 and GPT-3.5-Tu…

Hello Fellow Hackers, I wanted to share what my team is building. We released our open-source library for foundation model development in February and we're about to release our fi…

This is a really powerful technique in general because it lets us have some controllability over traditional PCG techniques! All you need is the right prompt and an evaluation metr…

On a related note, I've started a blog on procedural content generation and GenAI content synthesis: https://gamedev.blog/ . Would love any feedback / suggestions! I intend to cove…

There’s a similar library that also includes data synth and LLM-as-a-Judge: https://github.com/oumi-ai/oumi

Totally relate to that! Article looks interesting :)

There's quite a few differences between HuggingFace's Open Deep-Research and Zilliz's DeepSearcher. I think the biggest one is the goal: HF is to replicate the performance of Deep …

There's two blog posts that go with this, check it out: https://milvus.io/blog/i-built-a-deep-research-with-open-sou... https://milvus.io/blog/introduce-deepsearcher-a-local-open-s…