[WIP]: PromptTemplate

[EmanuelCampos]

DRAFT, not ready for use

missing:

• refactor missing modules
• refactor existent prompts
• tests with refine template
• fix circular dependency
• refactor partial format + prompt repack

[changeset-bot]

⚠️ No Changeset found

Latest commit: e658cfe

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
llama-index-ts-docs	❌ Failed (Inspect)			Mar 18, 2024 2:30pm

[EmanuelCampos]

this file will be removed, just an example to test fast for now

[marcusschiesser]

does this need to be function?

Suggested change

	const defaultTextQaPromptTemplate = () => `Context information is below.
	const defaultTextQaPromptTemplate = "Context information is below.

[marcusschiesser]

can we directly pass the template?

Suggested change

	"responseSynthesizer:textQATemplate": chatMessageTemplate,
	"responseSynthesizer:textQATemplate": [
	{
	content: "Always answer the question, even if you don't know the answer.",
	role: "system",
	},
	{
	content: defaultTextQaPromptTemplate(),
	role: "user",
	},
	],

e.g. updatePrompts could generate a ChatPromptTemplate object if needed

[marcusschiesser]

How about adding a convenience function for the user that is updating the most likely prompt (here responseSynthesizer:textQATemplate):

Suggested change

	queryEngine.updatePrompts({
	"responseSynthesizer:textQATemplate": chatMessageTemplate,
	queryEngine.updatePrompt(chatMessageTemplate)

[EmanuelCampos]

closing, due to several changes on how it works, moving to #665

@marcusschiesser thanks for the comments, appreciate your review on the new PR