Hyperion

Hyperion-shuo/picotron

Minimalistic 4D-parallelism distributed training framework for education purpose

Hyperion-shuo/lmdeploy

LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

Hyperion-shuo/xtuner

A Next-Generation Training Engine Built for Ultra-Large MoE Models

Hyperion-shuo/lagent

A lightweight framework for building LLM-based agents

Hyperion-shuo/defending-code-reference-harness

Skills for threat modeling, scanning, triage, patching, plus an autonomous scanning harness you can /customize

Hyperion-shuo/verl

verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

Hyperion-shuo/chasing_the_tail_rubrics

The official code repo of paper "Chasing the Tail: Effective Rubric-based Reward Modeling for Large Language Model Post-Training"

Hyperion-shuo/assignment5-alignment

Hyperion-shuo/assignment1-basics

Student version of Assignment 1 for Stanford CS336 - Language Modeling From Scratch

Hyperion-shuo/assignment2-systems

Student version of Assignment 2 for Stanford CS336 - Language Modeling From Scratch

Hyperion-shuo/assignment3-scaling

Hyperion-shuo/torch_dp_tutorial

Hyperion-shuo/LeetCode_Life

One leetcode everyday（python solution）

Hyperion-shuo/awr

Implementation of advantage-weighted regression.

Hyperion-shuo/Hyperion-shuo.github.io

Hyperion-shuo/OpenRLHF

An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & TIS & vLLM & Ray & Dynamic Sampling & Async Agentic RL)

Hyperion-shuo/slime

slime is an LLM post-training framework for RL Scaling.

Hyperion-shuo/reasoning-from-scratch

Implement a reasoning LLM in PyTorch from scratch, step by step

Hyperion-shuo/LLMs-from-scratch

Implement a ChatGPT-like LLM in PyTorch from scratch, step by step

Hyperion-shuo/nanochat

The best ChatGPT that $100 can buy.

Hyperion-shuo/reader3

Quick illustration of how one can easily read books together with LLMs. It's great and I highly recommend it.

Hyperion-shuo/nanoGPT

The simplest, fastest repository for training/finetuning medium-sized GPTs.

Hyperion-shuo/amago

off-policy RL on long sequences

Hyperion-shuo/clash-for-linux-install

😼 优雅地使用基于 clash/mihomo 的代理环境

Hyperion-shuo/agents-from-scratch

Build an email assistant with human-in-the-loop and memory

Hyperion-shuo/doraemon

Domain Randomization via Entropy Maximization

Hyperion-shuo/RetrievalTutorials

Hyperion-shuo/rag-from-scratch

Hyperion-shuo/minbpe

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.

Hyperion-shuo/example-code-2e

Example code for Fluent Python, 2nd edition (O'Reilly 2022)