Projects sharing features with Directional Preference Alignment

Arxiv

The Next Step Forward in Multimodal LLM Alignment

Aligning LMMs with Factually Augmented RLHF

This project is a transformer post-training toolkit and reinforcement learning library designed to align language model behavior with human preferences. It provides a framework for managing the transition from supervised fine-tuning to reinforcement learning and preference optimization. The library distinguishes itself through a specialized focus on preference optimization and reward modeling, enabling the adjustment of model outputs based on preferred versus rejected examples. It also includes capabilities for training agents within controlled sandbox environments using task suites and verif

Paper - Abstract - Updates - Quick Start - Installation - Download the datasets - Create Experiment Script - Single GPU Training (Only for Rho models) - Running the experiments - Code Structure - Initial SFT Checkpoints - Acknowledgement - Citation

DeepSpeedExamples is a collection of reference implementations for training and deploying large scale AI models using the DeepSpeed optimization library. It provides Python code examples for training massive models across multiple GPUs through distributed optimization techniques. The repository includes optimized patterns for deploying and running large language model predictions in production environments. It also serves as a guide for model compression to reduce memory footprints and as a source for performance benchmarks to measure execution speed and resource utilization. The project cov

Paper linkLINKTOPAPER

OpenRLHF is a training framework and alignment library designed for reinforcement learning from human feedback across distributed GPU clusters. It provides tools for aligning large language models and multimodal vision-language models using algorithms such as PPO, GRPO, and DPO. The framework distinguishes itself through a distributed inference engine that overlaps sample rollout with training to increase throughput. It supports scaling to models exceeding 70 billion parameters via parameter sharding and handles long-context sequences through ring-attention sequence parallelism. The project

LMFlow is a comprehensive suite for large language model fine-tuning, context extension, multimodal processing, and inference execution. It provides a toolkit for updating model parameters through full tuning or memory-efficient adapter algorithms, alongside an inference engine for executing tuned models via command-line or web-based interfaces. The framework includes a dedicated alignment suite for supervised tuning and reward model training to refine model behavior. It features a context window extender to increase maximum input lengths and a multimodal framework for building chatbots that

Align-anything is a multi-modal large language model alignment framework designed to fine-tune models across text, image, video, and audio. It functions as a distributed training orchestrator and toolkit for implementing preference-based learning to ensure model behaviors match human intentions and values. The framework provides specialized pipelines for Supervised Fine-Tuning and Direct Preference Optimization. It includes a high-performance inference engine wrapper for actor models to reduce sequence generation time and a dedicated training environment for refining vision-language-action mo

Safe RLHF: Constrained Value Alignment via Safe Reinforcement Learning from Human Feedback