A few weeks ago, OpenAI announced Reinforcement Fine-Tuning (RFT). This post covers the technical details of how it works, as well as the types of tasks for which it is a major breakthrough that lets LLMs be applied to even more complex custom tasks.

An engineer's "executive summary" on LLM fine-tuning

Chapter 2 continuing OpenPipe's fine-tuning best practices series. This chapter focuses specifically on models and the trade-off choices around them to ultimately create the highest performing fine-tuned LLM.

A post introducing a series of articles on LLM fine-tuning best practices, and the first chapter specifically on Training Data best practices.

We’re excited to announce a new family of “Mixture of Agents” models optimized for generating synthetic training data.

Here’s the summary and full video of a talk given at Mastering LLMs Conference titled the “Ten Commandments of Fine-Tuning in Prod.”

You’ve probably heard about Llama 3, a new open-source LLM. There are 3 variants announced so far: a small 8B parameter model, medium 70B parameter model, and very large 405B parameter model (still in training). I wanted to share the most important things we’ve learned about these models so far.

Hi, I’m David, the CTO of OpenPipe. This is a post for individuals who want to build high-level intuition around fine-tuning.

Today I’m excited to announce the close of our $6.7M seed round.

Evals can be tricky, but we’ve found that LLM-as-judge is the most reliable way to compare two outputs.

S-LoRA describes a set of optimizations for running thousands of separate LLMs simultaneously on the same GPU.

Axis is a modern platform to help multinational enterprises understand their local markets across the world.

Hi there! We'd love to share with you some of the recent updates we've made in December

Fine-tunes based on our new model are slightly stronger than GPT-4, as measured by GPT-4 itself.

At OpenPipe we’re building a fully-managed fine-tuning platform for developers.

Both crypto and AI have been heavily debated on Hacker News, with discussions going back years.

If you’re fine-tuning a task-specific LLM under 34B parameters you should probably be using Mistral.

I've been playing around with fine-tuning models for a couple of years, and wanted to share some insights and practical code

OpenPipe lets you capture your existing prompts and completions, and then use them to fine-tune a model specific to your use-case.