Every premium and members-only resource for $9.99/month, as a no-commitment subscription. See the plans

PalateKit

Sign in
Menu

Request a resource

Describe the resource you are missing in one sentence. We will email you when it is published.

0 / 280

A page, a video or an image online.

Unlock PalateKit

Unlock everything

Plan

Paying means accepting our terms of sale.

Secure payment by Stripe. Receipts and invoices sent by email.

Welcome back

Sign in to install, download or save this resource. A free account is enough for free resources.

PyTorch-native reinforcement learning with torchforge v1.0.0

Use this resource to experiment with reinforcement learning algorithms separated from infrastructure

New

  • 549 downloads
  • 130 likes
  • 110 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install torchforge-pytorch-rl-library

With no option, the skill goes to ~/.agents/skills/; add --agent claude-code (or codex, cursor…) to target an agent.

First time: connect the CLI with

npx --yes palatekit@latest login

The agent guides you through installing torchforge, checking the environment, then launching SFT training on two GPUs or GRPO on three GPUs. You get a setup where the algorithm is separated from infrastructure, with a ROCm installation guide.

Mostly picked by

Updated November 2, 2026

Card for the “PyTorch-native reinforcement learning with torchforge” skill

Details

Works in
Install guides
Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • SKILL.md: installation, architecture and SFT and GRPO runs
  • references/api-reference.md and troubleshooting.md: API and troubleshooting

Good fit for

  • You want to test your own reinforcement learning algorithms
  • You prefer a library with no Ray dependency

Not for

    Tags

    • pytorch
    • reinforcement
    • grpo
    • research

    About

    You say

    “Run a GRPO training with torchforge on three GPUs to compare two algorithm variants.”

    Card for the “PyTorch-native reinforcement learning with torchforge” skill

    Skill published by Orchestra Research under the MIT license, delivered in its original version.

    Show more

    How to use it

    1. Install the skill in your agent's skills folder
    2. Ask your agent to experiment with reinforcement learning algorithms separated from infrastructure
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/api-reference.md8.8 KB
    • references/troubleshooting.md7 KB
    • SKILL.md10.9 KB

    Origin and licence

    Listed unmodified

    Author
    Orchestra Research
    Source
    github.com/orchestra-research/ai-research-skills/tree/773a52944ba4747a18bd4ae9ade53fff041adcbc/06-post-training/torchforge
    Licence
    MIT

    Notice

    « torchforge-rl-training » (Orchestra Research, MIT) livré dans sa version d'origine : fiche en français et en anglais par PalateKit.

    User reviews

    –

    No ratings yet

    Loading reviews…

    Delete your review?

    Your rating and comment will be removed from the page.