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.

Supervised fine-tuning of vision-language models v1.0.0

Use this resource to fine-tune a vision-language model on image and text data

New

  • 414 downloads
  • 104 likes
  • 203 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install vision-language-model-finetuning

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 configures supervised fine-tuning of a VLM, with the vision tower and projector frozen and LoRA on the language model. It also helps diagnose a run that trains without the loss actually learning.

Mostly picked by

Updated November 2, 2026

Card for the “Supervised fine-tuning of vision-language models” skill

Details

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

Similar resources

View all

What you get

  • SKILL.md: core recipe and unfreezing rules
  • SKILL.md: failure modes and Unsloth specifics
  • references/collators-and-pitfalls.md: data collators and common pitfalls

Good fit for

  • You are adapting a VLM to a specific visual domain
  • Your fine-tune runs but the loss does not go down

Not for

    Tags

    • vision-language
    • fine-tuning
    • lora
    • image data

    About

    You say

    “Set up a LoRA fine-tune of my VLM on 2,000 image-caption pairs.”

    Card for the “Supervised fine-tuning of vision-language models” skill

    Skill published by wshobson 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 a vision-language model on image and text data
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/collators-and-pitfalls.md6 KB
    • SKILL.md7.9 KB

    Origin and licence

    Listed unmodified

    Author
    wshobson
    Source
    github.com/wshobson/agents/tree/0874a27b17a22e2b7c2e2bcd725768e0803cad06/plugins/llm-finetuning/skills/vision-sft
    Licence
    MIT

    Notice

    « vision-sft » (wshobson, 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.