An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.

RewardBench: the first evaluation tool for reward models.

Arena-Hard benchmark

Train neural networks up to 7x faster

Fast and flexible reference benchmarks

A modular RL library to fine-tune language models to human preferences

Essential guides and programming tools in my toolbox (with focus on ML Training)

A Data Streaming Library for Efficient Neural Network Training

NeMo: a toolkit for conversational AI

RLMeta is a light-weight flexible framework for Distributed Reinforcement Learning Research.

Implementation of the Off Belief Learning algorithm.

A reinforcement learning toolkit for compiler optimizations

A pytorch implementation of the paper "Prediction, Consistency, Curvature: Representation Learning for Locally-Linear Control"

Probabilistic reasoning and statistical analysis in TensorFlow

Computation using data flow graphs for scalable machine learning