How to Configure the Pieces for Sublime Extension

Supported LLMs

We constantly update and configure our plugins and extensions, like the Pieces for Sublime Text Plugin, to work with the latest LLMs.

Opening Pieces Settings

To open Pieces settings inside of the Pieces for Sublime Text Plugin, you can use ⌘+shift+p (macOS) or ctrl+shift+p (Windows/Linux) to open the command palette.

Then, enter Pieces: Open Pieces Settings.

Use the Pieces: Login or Pieces: Logout command in your command palette to log in and out of your Pieces account.

Connect / Disconnect your Personal Pieces Cloud

You can connect and disconnect from your personal cloud using Pieces: Connect Personal Cloud or Pieces: Disconnect Personal Cloud command in your command palette.

screenshot of Connect to Pieces Cloud command

Configurable Settings

The Pieces for Sublime Text Plugin offers customizable settings to tailor your experience. These options let you configure the connection, language model, startup behavior, and code autocompletion. Here's an overview of the key settings:

`host`	Specifies the address where the plugin should connect to a local service or server.
`model`	Determines which large language model to use for all Pieces Pilot functionality.
`auto_start_pieces_os`	Lets you decide whether or not to open PiecesOS automatically on startup.
`snippet.autocomplete`	Decide whether or not to enable auto-completion of code via suggest snippets from your repository.

To change settings, you need to copy and paste the contents of the default settings file into the empty window on the right side, which you can find below:

{
	"host": "",
	"model":"GPT-4o Chat Model",
			//  "GPT-4o Chat Model", 
			//  "GPT-4 Chat Model", 
			//  "CodeLlama GPU Chat Model",
			//  "GPT-3.5-turbo Chat Model",
			//  "Phi-2 CPU Chat Model", 
			//  "GPT-4 Preview Chat Model", 
			//  "Codey (PaLM2) Chat Model", 
			//  "Llama2 GPU Chat Model", 
			//  "(Gemini) Chat Model", 
			//  "GPT-3.5-turbo-16k Chat Model", 
			//  "Phi-2 GPU Chat Model", 
			//  "Mixtral GPU Chat Model", 
			//  "NeuralHermes-2.5-Mistral-7B CPU Chat Model", 
			//  "NeuralHermes-2.5-Mistral-7B GPU Chat Model", 
			//  "(PaLM2) Chat Model", 
			//  "Llama2 CPU Chat Model"
	"auto_start_pieces_os":false, // Open Pieces OS automatically on startup
	"snippet.autocomplete":true, // auto complete via Pieces saved snippets or not
	"syntax" : {} // Langauge extension : syntax path
}

Pieces Auto-Launch on Startup

You can choose whether or not PiecesOS starts immediately on the launch of the Sublime text editor by setting the parameters to true or false.

Snippet Auto-Complete

Similarly, adjust the parameters to true or false to enable auto-completion of code via snippets to speed up your coding progress.

If you have Pieces snippets saved that are relevant to the code you're writing, you'll get automatic completion for that specific language.

Reloading the Plugin

To reload the plugin, enter the command palette with ⌘+shift+p (macOS) or ctrl+shift+p (Windows / Linux) and type Pieces: Reload Plugin.

For additional support resources, check out our troubleshooting guide.

Saving Your Settings

Once you've copied the contents of the .JSON object and edited them to your preferences, simply save the file, reload the plugin, and restart Sublime for your changes to take effect.

You can save the .JSON file with ⌘+s (macOS) or ctrl+s (Windows / Linux). You do not need to use Save As, as the file is already existing and this is simply updating the settings file, which comes with the Pieces for Sublime Text package.

Configuration