---
title: RL
description: AI notes by Mrinaal Arora tagged "RL"
updated: 2026-04-29
canonical: https://aroramrinaal.com/ai/tags/rl
source: https://aroramrinaal.com/ai/tags/rl.md
alternate: https://aroramrinaal.com/ai/tags/rl/markdown
---

# RL

AI notes, small builds, and writeups by Mrinaal Arora tagged "RL".

- Page: https://aroramrinaal.com/ai/tags/rl
- Markdown: https://aroramrinaal.com/ai/tags/rl/markdown
- Markdown (.md): https://aroramrinaal.com/ai/tags/rl.md
- All tags: https://aroramrinaal.com/ai/tags/markdown
- Full AI index: https://aroramrinaal.com/ai/markdown
- Posts with this tag: 4 of 15

## Posts

| Post | Date | Tags | Word count | Source |
| --- | --- | --- | --- | --- |
| [CrisisOps: My (Final Round) OpenEnv Hackathon Project](https://aroramrinaal.com/ai/crisisops-openenv-hackathon-finale) | 2026-04-29 | RL | 1,767 words | [Markdown](https://aroramrinaal.com/ai/crisisops-openenv-hackathon-finale/markdown) |
| [My First Successful RL Training Run With OpenEnv And Wordle](https://aroramrinaal.com/ai/rl-training-openenv-environment) | 2026-04-02 | RL | 2,095 words | [Markdown](https://aroramrinaal.com/ai/rl-training-openenv-environment/markdown) |
| [My Second RL Experiment On Prime Intellect Lab](https://aroramrinaal.com/ai/second-rl-experiment-on-prime-intellect-lab) | 2026-03-25 | RL | 1,984 words | [Markdown](https://aroramrinaal.com/ai/second-rl-experiment-on-prime-intellect-lab/markdown) |
| [First RL Experimental Project: RLVR Using GRPO With TRL On Modal](https://aroramrinaal.com/ai/first-rl-experimental-project) | 2026-03-23 | RL | 1,856 words | [Markdown](https://aroramrinaal.com/ai/first-rl-experimental-project/markdown) |

## CrisisOps: My (Final Round) OpenEnv Hackathon Project

- Slug: crisisops-openenv-hackathon-finale
- Date: 2026-04-29
- Tags: RL
- Word count: 1,767 words
- Page: https://aroramrinaal.com/ai/crisisops-openenv-hackathon-finale
- Markdown: https://aroramrinaal.com/ai/crisisops-openenv-hackathon-finale/markdown
- Markdown (.md): https://aroramrinaal.com/ai/crisisops-openenv-hackathon-finale.md
- X: https://x.com/arora_mrinaal/status/2047992775162138805?s=20
- Hugging Face: https://huggingface.co/spaces/mrinaalarora/crisisops

How I built CrisisOps for the OpenEnv hackathon finale, then trained a small model with GRPO using TRL, Unsloth, and Modal.
## My First Successful RL Training Run With OpenEnv And Wordle

- Slug: rl-training-openenv-environment
- Date: 2026-04-02
- Tags: RL
- Word count: 2,095 words
- Page: https://aroramrinaal.com/ai/rl-training-openenv-environment
- Markdown: https://aroramrinaal.com/ai/rl-training-openenv-environment/markdown
- Markdown (.md): https://aroramrinaal.com/ai/rl-training-openenv-environment.md
- X: https://x.com/arora_mrinaal/status/2038914573227667505?s=20
- Hugging Face: https://huggingface.co/mrinaalarora/wordle-grpo-Qwen3-1.7B

A full fine-tune of Qwen3-1.7B on Wordle using OpenEnv, with a reward curve that actually went up and shaped rewards that taught real gameplay behavior.
## My Second RL Experiment On Prime Intellect Lab

- Slug: second-rl-experiment-on-prime-intellect-lab
- Date: 2026-03-25
- Tags: RL
- Word count: 1,984 words
- Page: https://aroramrinaal.com/ai/second-rl-experiment-on-prime-intellect-lab
- Markdown: https://aroramrinaal.com/ai/second-rl-experiment-on-prime-intellect-lab/markdown
- Markdown (.md): https://aroramrinaal.com/ai/second-rl-experiment-on-prime-intellect-lab.md
- X: https://x.com/arora_mrinaal/status/2036405167701033343

My second RL experiment while studying GRPO: moving to Prime Intellect Lab, testing smaller and larger setups, and finally getting a reward curve that went up.
## First RL Experimental Project: RLVR Using GRPO With TRL On Modal

- Slug: first-rl-experimental-project
- Date: 2026-03-23
- Tags: RL
- Word count: 1,856 words
- Page: https://aroramrinaal.com/ai/first-rl-experimental-project
- Markdown: https://aroramrinaal.com/ai/first-rl-experimental-project/markdown
- Markdown (.md): https://aroramrinaal.com/ai/first-rl-experimental-project.md
- X: https://x.com/arora_mrinaal/status/2035385790910374136

My first ever RL experiment: RLVR on GSM8K using Hugging Face TRL, Qwen2.5 1.5B Instruct, and an NVIDIA H100 on Modal, with notes on where the reward function broke and what I want to fix in v2.