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.

Fine-tuning and aligning LLMs with TRL v1.0.0

Use this resource to fine-tune an LLM with SFT, DPO or GRPO and train a reward model

New

  • 579 downloads
  • 106 likes
  • 229 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install trl-sft-dpo-rlhf-training

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 TRL, running supervised fine-tuning, aligning the model on preference pairs with DPO, then training a reward model for a full RLHF pipeline. It builds on Transformers.

Mostly picked by

Updated November 2, 2026

Card for the “Fine-tuning and aligning LLMs with TRL” skill

Details

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

Similar resources

View all

What you get

  • SKILL.md: installation, SFT, DPO and a basic RLHF pipeline
  • references/sft-training.md, dpo-variants.md, reward-modeling.md and online-rl.md: SFT, DPO variants, reward models and online RL

Good fit for

  • You want to align a model on your users' preferences
  • You build a full RLHF pipeline with Hugging Face

Not for

    Tags

    • rlhf
    • dpo
    • sft
    • alignment

    About

    You say

    “Run SFT then DPO on Qwen 0.5B using my preferred response pairs.”

    Card for the “Fine-tuning and aligning LLMs with TRL” 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 fine-tune an LLM with SFT, DPO or GRPO and train a reward model
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/dpo-variants.md4.4 KB
    • references/online-rl.md2 KB
    • references/reward-modeling.md2.7 KB
    • references/sft-training.md3.3 KB
    • SKILL.md11.6 KB

    Origin and licence

    Listed unmodified

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

    Notice

    « fine-tuning-with-trl » (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.