Training Sparse Autoencoders on Language Models

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

MiniCPM-2B: An end-side LLM outperforming Llama2-13B.

A WebUI for Efficient Fine-Tuning of 100+ LLMs (ACL 2024)

HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal

Go ahead and axolotl questions

The TinyLlama project is an open endeavor to pretrain a 1.1B Llama model on 3 trillion tokens.

Collection of papers for scalable automated alignment.

RewardBench: the first evaluation tool for reward models.

An educational resource to help anyone learn deep reinforcement learning.

Scalable toolkit for efficient model alignment

For OpenMOSS Mechanistic Interpretability Team's Sparse Autoencoder (SAE) research.

SimPO: Simple Preference Optimization with a Reference-Free Reward

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.

A series of large language models trained from scratch by developers @01-ai

Code release for "Debating with More Persuasive LLMs Leads to More Truthful Answers"

Paper list of multi-agent reinforcement learning (MARL)

Self-playing Adversarial Language Game Enhances LLM Reasoning

A JAX research toolkit for building, editing, and visualizing neural networks.

This is a collection of research papers for Self-Correcting Large Language Models with Automated Feedback.

[ACL 2024] A Survey of Chain of Thought Reasoning: Advances, Frontiers and Future

A curated list of reinforcement learning with human feedback resources (continually updated)

Qwen2 is the large language model series developed by Qwen team, Alibaba Cloud.

Easy-to-Hard Generalization: Scalable Alignment Beyond Human Supervision

Simple and efficient pytorch-native transformer training and inference (batched)