Add post about Web Retrieval #242
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
|
|
||
| # Searching the Web | ||
|
|
||
| As we've seen [earlier](#), a Haystack RAG Pipeline is made of three components: a Retriever, a PromptBuilder, and a Generator, and looks like this: |
There was a problem hiding this comment.
This was linking to a post that we haven't published here yet. LMK what to point it to.
There was a problem hiding this comment.
just remove the first bit omi, Start with 'A Haystack RAG pipeline...'
|
|
||
| # Processing the page | ||
|
|
||
| In a [previous post](#), we've seen how Haystack can convert web pages into clean Documents ready to be stored in a Document Store. We will reuse many of the components we have discussed there, so if you missed it, make sure to check it out. |
There was a problem hiding this comment.
This was linking to a post that we haven't published here yet. LMK what to point it to.
There was a problem hiding this comment.
Again, same as previous comment, we will probably publish this first so we can remove this line?
There was a problem hiding this comment.
I've added some comments, @bilgeyucel might have more. Also, do we have to change the headers to h2? Not sure.
| tags: ["Haystack 2.0"] | ||
| --- | ||
|
|
||
| In an earlier post of the Haystack 2.0 series, we've seen how to build RAG and indexing pipelines. An application that uses these two pipelines is practical if you have an extensive, private collection of documents and need to perform RAG on such data only. However, in many cases, you may want to get data from the Internet: from news outlets, documentation pages, and so on. |
There was a problem hiding this comment.
I would rephrase because we will probably post this before the others: "An application that uses these two pipelines is practical if you have an extensive, private collection of documents and need to perform RAG on such data only. One pipeline to index documents to your own, private document store (dubbed an indexing pipeline), and another to query it with a language model (the querying pipeline). However,..."
|
|
||
| # Searching the Web | ||
|
|
||
| As we've seen [earlier](#), a Haystack RAG Pipeline is made of three components: a Retriever, a PromptBuilder, and a Generator, and looks like this: |
There was a problem hiding this comment.
just remove the first bit omi, Start with 'A Haystack RAG pipeline...'
|
|
||
| # Processing the page | ||
|
|
||
| In a [previous post](#), we've seen how Haystack can convert web pages into clean Documents ready to be stored in a Document Store. We will reuse many of the components we have discussed there, so if you missed it, make sure to check it out. |
There was a problem hiding this comment.
Again, same as previous comment, we will probably publish this first so we can remove this line?
The thumbnail is from Wikipedia, copyright free. Let me know if and how I should provide attribution.