GRPO training code which scales to 32xH100s for long horizon terminal/coding tasks. Base agent is now the top Qwen3 agent on Stanford's TerminalBench leaderboard.
☆412Aug 24, 2025Updated last year
Alternatives and similar repositories for terminal-bench-rl
Users that are interested in terminal-bench-rl are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Multi-agent synthetic data generation pipeline capable of generating and validating long horizon terminal/coding tasks for RL training☆74Jul 28, 2025Updated last year
- 💻 SETA: Scaling Environments for Terminal Agents☆156Jul 28, 2026Updated last month
- A benchmark for LLMs on complicated tasks in the terminal☆2,593Jul 11, 2026Updated 2 months ago
- 💻 SETA: Scaling Environments for Terminal Agents - Environments☆147Feb 16, 2026Updated 7 months ago
- Democratizing Reinforcement Learning for LLMs☆5,836Sep 12, 2026Updated last week
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Data recipes and robust infrastructure for training AI agents☆295Updated this week
- Code for Paper: Training Software Engineering Agents and Verifiers with SWE-Gym [ICML 2025]☆745Jul 29, 2025Updated last year
- ☆148Mar 31, 2026Updated 5 months ago
- SkyRL: A Modular Full-stack RL Library for LLMs☆2,346Updated this week
- A Gym for Agentic LLMs☆510Jan 21, 2026Updated 8 months ago
- Convert GitHub PRs into Harbor tasks☆85Jul 13, 2026Updated 2 months ago
- [NeurIPS 2025 D&B Spotlight] Scaling Data for SWE-agents☆786Updated this week
- Framework for evaluating and improving agents☆5,571Updated this week
- ☆39Apr 2, 2026Updated 5 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- [COLM 2025] Official repository for R2E-Gym: Procedural Environment Generation and Hybrid Verifiers for Scaling Open-Weights SWE Agents☆335Jul 13, 2025Updated last year
- slime is an LLM post-training framework for RL Scaling.☆8,527Updated this week
- Our library for RL environments + evals☆4,650Updated this week
- Understanding R1-Zero-Like Training: A Critical Perspective☆1,278Aug 27, 2025Updated last year
- [ICLR 2026] The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution☆489Aug 18, 2026Updated last month
- Agentic RL Training at Scale☆2,081Updated this week
- ☆123Apr 1, 2026Updated 5 months ago
- [ICLR 2026] End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoning☆405Mar 30, 2026Updated 5 months ago
- ☆15Oct 13, 2025Updated 11 months ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- ☆17Jan 27, 2026Updated 7 months ago
- [NeurIPS'25] Official codebase for "SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution"☆720Mar 16, 2025Updated last year
- ☆415Apr 30, 2026Updated 4 months ago
- ☆15Jan 14, 2026Updated 8 months ago
- The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.☆5,793Updated this week
- Sandboxed code execution for AI agents, locally or on the cloud. Massively parallel, easy to extend. Powering SWE-agent and more.☆601Updated this week
- verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework☆23,601Updated this week
- An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Asy…☆10,042Sep 17, 2026Updated last week
- A version of verl to support diverse tool use [TMLR 2026]☆1,044Jul 15, 2026Updated 2 months ago
- AI Agents on DigitalOcean Gradient AI Platform • AdBuild production-ready AI agents using customizable tools or access multiple LLMs through a single endpoint. Create custom knowledge bases or connect external data.
- Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.☆2,979Updated this week
- ☆44May 16, 2026Updated 4 months ago
- [NeurIPS 2025 Spotlight] Reasoning Environments for Reinforcement Learning with Verifiable Rewards☆1,514Apr 17, 2026Updated 5 months ago
- A technical exploration comparing standard deep RL (PPO) against biologically plausible learning rules on a custom Pong environment. Ever…☆29May 19, 2026Updated 4 months ago
- Container-free RL framework for training software engineering agents☆79Jun 24, 2026Updated 3 months ago
- Official Repo for Open-Reasoner-Zero☆2,098Jun 2, 2025Updated last year
- ☆177May 13, 2026Updated 4 months ago