Fine-Tuning Workspace

Fine-Tuning Workspace is a VS Code extension that scaffolds a repeatable workspace for LLM fine-tuning projects. It creates a ready-to-organize folder layout for datasets, LoRA adapters, training configs, evaluation assets, deployment files, and project documentation.

Features

• Adds the Create Fine-Tuning Workspace command to VS Code.
• Creates a complete directory structure for data preparation, training, evaluation, deployment, and docs.
• Copies starter templates for FastAPI fine-tuning workflows into the generated workspace.
• Gives teams a consistent project shape so experiments are easier to manage and share.

Command

Run the command palette action below:

• Create Fine-Tuning Workspace

The extension prompts for a target folder, then generates a workspace with folders such as:

• datasets/
• loras/
• models/
• training/
• evaluation/
• deployment/
• docs/

It also adds starter files including:

• training/scripts/train_fastapi_lora.py
• training/configs/fastapi.yaml
• docs/fastapi-lora-card.md

Requirements

There are no extension-side configuration requirements. You only need VS Code 1.120.0 or newer.

The generated workspace may include Python-oriented templates for model training, so your runtime requirements depend on how you use those generated files.

Extension Settings

This extension does not currently contribute any custom VS Code settings.

Known Issues

• The generated scaffolding is intentionally opinionated and currently focused on a FastAPI-oriented template set.
• Packaging may still warn about explicit activationEvents because the command contribution is enough for modern VS Code.

Release Notes

0.0.1

Initial release with workspace scaffolding, training templates, and extension icon support.