Prime Intellect ORG

Our library for RL environments + evals

Agentic RL Training at Scale

Official CLI and Python SDK for Prime Intellect - access GPU compute, remote sandboxes, RL environments, and distributed training infrastructure for AI development at scale.

Programmable chat templates for LLM training and inference.

Environments by the Prime Intellect Research Team

Lightly-reviewed collection of community environments

prime is a framework for efficient, globally distributed training of AI models over the internet.

OpenDiLoCo: An Open-Source Framework for Globally Distributed Low-Communication Training

PCCL (Prime Collective Communications Library) implements fault tolerant collective communications over IP

peer-to-peer compute and intelligence network that enables decentralized AI development at scale

No description.

autonomous nanogpt optimizer speedrun

SIMD quantization kernels

TOPLOC: is a novel method for verifiable inference that enables users to verify that LLM providers are using the correct model configurations and settings

Asynchronous P2P communication backend for decentralized pipeline parallelism

A 7B parameter model for mathematical reasoning

Modded vLLM to run pipeline parallelism over public networks

Lab Cookbook

Solidity contracts for the decentralized Prime Network protocol

torch implementation of diloco

Harness for RLM-style rollouts. Only for RL training

C99-compatible library for efficiently parking threads on all major operating systems

Research sandbox for decentralized pipelined inference

Cryptographically verifiable GPU matmul challenge

Decentralized deep learning in PyTorch. Built to train models on thousands of volunteers across the world.

No description.

Fast & lightweight C++ threadpool

No description.

No description.

No description.

SGLang is a fast serving framework for large language models and vision language models.

Specify what you want it to build, the AI asks for clarification, and then builds it.

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

No description.

Automatic evals for LLMs

Large Context Transformers

Democratizing Reinforcement Learning for LLMs

The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery 🧑‍🔬

Large-scale Harbor task data backing research-environments tasksets.

Manage ML configuration with pydantic

MCP Environments demo

A high-performance and light-weight router for vLLM large scale deployment

Code for Parsel 🐍 - generate complex programs with language models

No description.

The open source coding agent.

PyTorch RFCs (experimental)

For optimization algorithm research and development.

A simple performant dataset streaming server & client for tokenized webtext datasets

No description.

A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch

No description.

C++ library for solving the traveling salesman problem efficiently

No description.

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

code example for prime platform documentation

No description.

System 2 Reasoning Link Collection

LlamaIndex (formerly GPT Index) is a data framework for your LLM applications

No description.