Recipes to train reward model for RLHF.

[NeurIPS 2023] Reflexion: Language Agents with Verbal Reinforcement Learning

Official implementation of project Honeybee (CVPR 2024)

A very simple performing matrix multiplication example for CPU / CUDA / METAL using GGML / llama.cpp

LLM101n: Let's build a Storyteller

Efficiently computes derivatives of NumPy code.

The official implementation of “Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training”

Implemented the forward mode of automatic differentiation with the help of dual numbers using Python.

Pure Python autograd library based on NumPy

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

Flash Attention in ~100 lines of CUDA (forward pass only)

Large scale 4D parallelism pre-training for 🤗 transformers in Mixture of Experts *(still work in progress)*

A high-throughput and memory-efficient inference and serving engine for LLMs

A native PyTorch Library for large model training

The official Meta Llama 3 GitHub site

Regex for Korean curseword filtering

symspellpy를 한글 특성에 맞춰서 수정한 라이브러리. 음소분해를 이용해 더 정확한 오타교정을 해준다.

Repo for Rho-1: Token-level Data Selection & Selective Pretraining of LLMs.

Scalable data pre processing and curation toolkit for LLMs

KoLLaVA: Korean Large Language-and-Vision Assistant (feat.LLaVA)

Collection of leaked system prompts

Implementation of MEGABYTE, Predicting Million-byte Sequences with Multiscale Transformers, in Pytorch

Minimalistic large language model 3D-parallelism training