Skip to main content

Openrlhf Training

Skill Verified Active

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

Purpose

To enable efficient and scalable training of large language models using advanced RLHF techniques with high-performance distributed architecture.

Features

  • High-performance RLHF training
  • Ray + vLLM acceleration
  • Support for PPO, GRPO, RLOO, DPO
  • Distributed architecture for large models
  • GPU resource sharing via Hybrid Engine

Use Cases

  • Training large language models (7B-70B+) with RLHF
  • Achieving 2x faster training compared to DeepSpeedChat
  • Leveraging multi-node GPU clusters for distributed training
  • Fine-tuning models with advanced RL algorithms in a unified framework

Non-Goals

  • General-purpose model fine-tuning outside of RLHF
  • Inference serving or deployment orchestration
  • Model architecture definition or modification

Installation

First, add the marketplace

/plugin marketplace add Orchestra-Research/AI-Research-SKILLs
/plugin install AI-Research-SKILLs@ai-research-skills

Quality Score

Verified
99 /100
Analyzed about 20 hours ago

Trust Signals

Last commit16 days ago
Stars8.3k
LicenseMIT
Status
View Source

Similar Extensions

Ray Train

99

Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling. Use when training massive models across multiple machines or running distributed hyperparameter sweeps.

Skill
Orchestra-Research

Openrlhf Training

97

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

Skill
davila7

Pytorch Lightning

99

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

Skill
Orchestra-Research

Verl Rl Training

99

Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

Skill
Orchestra-Research

TorchTitan Distributed LLM Pretraining

99

Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

Skill
Orchestra-Research

Huggingface Accelerate

99

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.

Skill
davila7

© 2025 SkillRepo · Find the right skill, skip the noise.