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.
☆419Aug 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☆76Jul 28, 2025Updated last year
- A benchmark for LLMs on complicated tasks in the terminal☆2,613Jul 11, 2026Updated 3 months ago
- 💻 SETA: Scaling Environments for Terminal Agents☆197Sep 30, 2026Updated last month
- Democratizing Reinforcement Learning for LLMs☆5,910Oct 16, 2026Updated 2 weeks ago
- 💻 SETA: Scaling Environments for Terminal Agents - Environments☆152Feb 16, 2026Updated 8 months ago
- GPUs on demand by Runpod - Special Offer Available • AdRun AI, ML, and HPC workloads on powerful cloud GPUs—without limits or wasted spend. Deploy GPUs in under a minute and pay by the second.
- Code for Paper: Training Software Engineering Agents and Verifiers with SWE-Gym [ICML 2025]☆757Updated this week
- SkyRL: A Modular Full-stack RL Library for LLMs☆2,498Updated this week
- A Gym for Agentic LLMs☆514Jan 21, 2026Updated 9 months ago
- ☆148Mar 31, 2026Updated 7 months ago
- [NeurIPS 2025 D&B Spotlight] Scaling Data for SWE-agents☆823Updated this week
- Data recipes and robust infrastructure for training AI agents☆309Oct 23, 2026Updated last week
- ☆76Mar 24, 2026Updated 7 months ago
- Convert GitHub PRs into Harbor tasks☆89Jul 13, 2026Updated 3 months ago
- ☆128Apr 1, 2026Updated 7 months ago
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- Training terminal-agents☆340Updated this week
- A home for run logs for terminal-bench☆15Nov 14, 2025Updated 11 months ago
- slime is an LLM training framework for RL Scaling.☆8,794Updated this week
- Understanding R1-Zero-Like Training: A Critical Perspective☆1,287Aug 27, 2025Updated last year
- ☆41Apr 2, 2026Updated 7 months ago
- [COLM 2025] Official repository for R2E-Gym: Procedural Environment Generation and Hybrid Verifiers for Scaling Open-Weights SWE Agents☆347Jul 13, 2025Updated last year
- Our library for RL environments + evals☆4,768Updated this week
- This repository, deepspeed-grpo-qlora-vllm, provides a complete framework for fine-tuning LLMs using Group Relative Policy Optimization (…☆13Jul 15, 2025Updated last year
- Framework for evaluating and improving agents☆6,821Updated this week
- GPUs on demand by Runpod - Special Offer Available • AdRun AI, ML, and HPC workloads on powerful cloud GPUs—without limits or wasted spend. Deploy GPUs in under a minute and pay by the second.
- [NeurIPS'25] Official codebase for "SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution"☆721Mar 16, 2025Updated last year
- ☆18Sep 11, 2025Updated last year
- ☆441Apr 30, 2026Updated 6 months ago
- ☆46May 16, 2026Updated 5 months ago
- [ICLR 2026] The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution☆513Aug 18, 2026Updated 2 months ago
- [ICLR 2026] End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated Reasoning☆408Mar 30, 2026Updated 7 months ago
- Agentic RL Training at Scale☆2,218Updated this week
- Terminal-Lego: What Makes Interaction Trajectories Effective for Training Terminal Agents?☆13Jun 4, 2026Updated 5 months ago
- The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.☆5,901Updated this week
- Managed Database hosting by DigitalOcean • AdPostgreSQL, MySQL, MongoDB, Kafka, Valkey, and OpenSearch available. Automatically scale up storage and focus on building your apps.
- Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.☆3,272Updated this week
- [NeurIPS 2025 Spotlight] Reasoning Environments for Reinforcement Learning with Verifiable Rewards☆1,545Apr 17, 2026Updated 6 months ago
- ☆15Oct 13, 2025Updated last year
- ☆17Jan 27, 2026Updated 9 months ago
- verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework☆24,143Updated this week
- Sandboxed code execution for AI agents, locally or on the cloud. Massively parallel, easy to extend. Powering SWE-agent and more.☆632Updated this week
- Measuring and evolving with the frontier of agent work☆1,013Updated this week