How do I install the Justrl Simple Recipe skill?

Run npx skillmds@latest add adu2021/justrl-simple-recipe in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Justrl Simple Recipe safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Justrl Simple Recipe?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Justrl Simple Recipe free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.

Who published Justrl Simple Recipe?

adu2021 (@adu2021) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Justrl Simple Recipe by adu2021 — Agent Skill

Name: Justrl Simple Recipe
Author: adu2021

Justrl Simple Recipe

Demonstrate that simple single-stage RL with fixed hyperparameters matches complex multi-stage approaches for training small LLMs on mathematical reasoning. Use basic setup: GRPO algorithm, rule-based verification, 16K token context, standard training data without difficulty filtering. Train two 1.5B models to competitive performance using 2× less compute than sophisticated approaches.

adu2021 d82da62 3.5 KB Updated Sep 17, 2026

File contents

Skill Summary

JustRL challenges prevailing assumptions about LLM training complexity. The paper demonstrates that simple, single-stage reinforcement learning with fixed hyperparameters can match or exceed complex multi-stage approaches for training small language models on mathematical reasoning. The minimal recipe uses GRPO algorithm with rule-based verification, 16K token context limit, and standard training data without difficulty filtering. Training dynamics exhibit smooth monotonic improvement without collapses or plateaus typically motivating multi-stage interventions. Two 1.5B models (DeepSeek and Nemotron backbones) achieve competitive performance (54.9% and 64.3% average accuracy respectively) while using 2× less compute.

When To Use

Training small language models (1-3B) on mathematical reasoning
Projects skeptical of complex multi-stage RL schemes
Scenarios where simple approaches reduce infrastructure complexity
Research questioning necessity of sophisticated training techniques

When NOT To Use

Large-scale model training (findings may not scale beyond 1.5B)
Tasks benefiting specifically from curriculum learning or multi-stage approaches
Scenarios where simple methods genuinely underperform
Domains requiring complex RL dynamics beyond basic GRPO

Core Technique

The recipe intentionally minimizes complexity:

1. Single-Stage Training No progressive context lengthening, no curriculum switching, no dynamic adjustments. Apply same training approach uniformly throughout.

2. Fixed Hyperparameters Use identical hyperparameters across all models and training stages. No adaptive scheduling or dynamic adjustments based on training progress.

3. Standard Training Data Use basic training data without difficulty filtering, dynamic sampling, or curriculum-based selection. Standard dataset for baseline performance.

4. Basic Algorithm

GRPO: standard group policy optimization
Rule-based verification: simple correctness checking
16K token context: moderate but fixed context window

Key Insight

As authors note, the approach achieves results "while using 2× less compute than sophisticated approaches." Training dynamics exhibit smooth, monotonic improvement without collapses or plateaus typically motivating multi-stage interventions. This suggests that "adequate scale with simple methods" may suffice where the field has added unnecessary complexity.

Results

DeepSeek 1.5B: 54.9% average accuracy across nine benchmarks
Nemotron 1.5B: 64.3% average accuracy
Smooth training without catastrophic forgetting
2× compute efficiency vs. multi-stage approaches

Implementation Notes

Implement basic GRPO algorithm with fixed hyperparameters. Use rule-based verification for mathematical correctness. Train with standard data without curriculum or difficulty filtering. Monitor for smooth improvement trajectory. Validate that simple approaches work before adding complexity.

References

Original paper: JustRL (Dec 2025)
Group Policy Optimization (GRPO)
Language model training recipes

---name: justrl-simple-recipedescription: Demonstrate that simple single-stage RL with fixed hyperparameters matches complex multi-stage approaches for training small LLMs on mathematical reasoning. Use basic setup: GRPO algorithm, rule-based verification, 16K token context, standard training data without difficulty filtering. Train two 1.5B models to competitive performance using 2× less compute than sophisticated approaches.license: MIT---## Skill SummaryJustRL challenges prevailing assumptions about LLM training complexity. The paper demonstrates that simple, single-stage reinforcement learning with fixed hyperparameters can match or exceed complex multi-stage approaches for training small language models on mathematical reasoning. The minimal recipe uses GRPO algorithm with rule-based verification, 16K token context limit, and standard training data without difficulty filtering. Training dynamics exhibit smooth monotonic improvement without collapses or plateaus typically motivating multi-stage interventions. Two 1.5B models (DeepSeek and Nemotron backbones) achieve competitive performance (54.9% and 64.3% average accuracy respectively) while using 2× less compute.## When To Use- Training small language models (1-3B) on mathematical reasoning- Projects skeptical of complex multi-stage RL schemes- Scenarios where simple approaches reduce infrastructure complexity- Research questioning necessity of sophisticated training techniques## When NOT To Use- Large-scale model training (findings may not scale beyond 1.5B)- Tasks benefiting specifically from curriculum learning or multi-stage approaches- Scenarios where simple methods genuinely underperform- Domains requiring complex RL dynamics beyond basic GRPO## Core TechniqueThe recipe intentionally minimizes complexity:**1. Single-Stage Training**No progressive context lengthening, no curriculum switching, no dynamic adjustments. Apply same training approach uniformly throughout.**2. Fixed Hyperparameters**Use identical hyperparameters across all models and training stages. No adaptive scheduling or dynamic adjustments based on training progress.**3. Standard Training Data**Use basic training data without difficulty filtering, dynamic sampling, or curriculum-based selection. Standard dataset for baseline performance.**4. Basic Algorithm**- GRPO: standard group policy optimization- Rule-based verification: simple correctness checking- 16K token context: moderate but fixed context window## Key InsightAs authors note, the approach achieves results "while using 2× less compute than sophisticated approaches." Training dynamics exhibit smooth, monotonic improvement without collapses or plateaus typically motivating multi-stage interventions. This suggests that "adequate scale with simple methods" may suffice where the field has added unnecessary complexity.## Results- DeepSeek 1.5B: 54.9% average accuracy across nine benchmarks- Nemotron 1.5B: 64.3% average accuracy- Smooth training without catastrophic forgetting- 2× compute efficiency vs. multi-stage approaches## Implementation NotesImplement basic GRPO algorithm with fixed hyperparameters. Use rule-based verification for mathematical correctness. Train with standard data without curriculum or difficulty filtering. Monitor for smooth improvement trajectory. Validate that simple approaches work before adding complexity.## References- Original paper: JustRL (Dec 2025)- Group Policy Optimization (GRPO)- Language model training recipes

Frequently asked questions