stefanwebb

5 ms·

1.
Show HN: A better alternative to CLI and MCP for local tools(github.com)
2 points by stefanwebb 6mo ago|0 comments
2.
Open source x 3: GRPO training with OpenEnv, vLLM, and Oumi(github.com)
3 points by stefanwebb 11mo ago|0 comments
3.
Is more training data always better?(blog.oumi.ai)
3 points by stefanwebb 1y ago|2 comments
4.
by stefanwebb 1y ago
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], attention-based neural networks [Vaswani et al., 2017], and
5.
by stefanwebb 1y ago
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 to outperform GPT-4 called "
6.
Small Fine-Tuned Models Are All You Need(blog.oumi.ai)
6 points by stefanwebb 1y ago|2 comments
7.
by stefanwebb 1y ago
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-Turbo, as well as arguments why fine-tuning is coming back int
8.
Custom AI models in hours not months with auto Data Synth and LLM-as-a-Judge(blog.oumi.ai)
3 points by stefanwebb 1y ago|1 comments
9.
by stefanwebb 1y ago
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 first Enterprise offering. In brief, we've developed
10.
by stefanwebb 1y ago
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 metric - could definitely apply to Voronoi maps
11.
by stefanwebb 1y ago
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 cover Voronoi diagrams in the near fut
12.
by stefanwebb 1y ago
There’s a similar library that also includes data synth and LLM-as-a-Judge: https://github.com/oumi-ai/oumi
13.
by stefanwebb 2y ago
Totally relate to that! Article looks interesting :)
14.
by stefanwebb 2y ago
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 Research on the GAIA benchmark whereas ours i
15.
by stefanwebb 2y ago
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...