alibaba/ROLL
ROLL
An Efficient and User-Friendly Scaling Library for Reinforcement Learning with Large Language Models
Builder
alibaba
alibaba • individual
Stars
3,042
Using upstream star count
Forks
262
Using upstream fork count
Open Issues
0
Activity Score
0/100
0 commits in 30d
Created
May 28, 2025
Project creation date
README Summary
ROLL is a scaling library that enables efficient distributed training and inference of reinforcement learning algorithms combined with large language models. It provides user-friendly APIs and tools for researchers and practitioners to scale RL-LLM systems across multiple GPUs and nodes.
AI Dev Skills
Unmapped
Reinforcement LearningLarge Language Model TrainingDistributed ComputingModel ParallelismGPU OptimizationDeep Reinforcement LearningLanguage Model Fine-tuningScalable ML SystemsParallel Processing
Tags
Reinforcement LearningLarge Language Model TrainingDistributed ComputingModel ParallelismGPU OptimizationDeep Reinforcement LearningLanguage Model Fine-tuningScalable ML SystemsParallel Processing
Taxonomy
AI Trends
Deployment Context
Modalities
Skill Areas
Recent Activity
Updated 16 days ago
7 Days
0
30 Days
0
90 Days
0
Quality
research- Quality
- medium
- Maturity
- research
Categories
Dev Tools & AutomationPrimaryFoundation ModelsModel TrainingInference & ServingOther AI / ML
PM Skills
Developer Platform
Languages
Python100.0%
Timeline
- Project created
- May 28, 2025
- Forked
- Mar 29, 2026
- Your last push
- 16 days ago
- Upstream last push
- 7 days ago
- Tracked since
- Mar 29, 2026
Similar Repos
pgvector cosine similarity · $0
Loading…