How do I retrieve my embeddings

Hi, beginner question.

I have use this tutorial effectively to create my first pinecone index with a bunch of PDFs I have

(1) LangChain101: Question A 300 Page Book (w/ OpenAI + Pinecone) - YouTube

I now want to (1) LangChain101: Question A 300 Page Book (w/ OpenAI + Pinecone) - YouTube

Now when I restart my application I dont want to create new embeddings again for existing docs but want to retrieve all the embeddings from the documents in my index.

How do I do this? ( have searched around and even asked GPT-4 but went down a rabbit hole!)

Thanks
JK

Ok will close out my own request, seems others have struggled with this.

Query on existing index · Issue #1792 · hwchase17/langchain (github.com)

Answer is:


embeddings = OpenAIEmbeddings()
docsearch = Pinecone.from_existing_index(index_name=index_name, embedding=embeddings, namespace="your_namespace_name")

no embeddings are not coming by this way. the way you explained. I have multiple documents in same namespace. How do i extract the embeddings of few specific.

Were you able to get a solution? I am also confused that how do I get embeddings of relevant documents only.

Were you able to get a solution? I am also confused that how do I get embeddings of relevant documents only.

Its been a year, why is there no response from the pincone team on this?

Not sure if anyone was able to help out, but I just made it work with:

pinecone = PineconeVectorStore.from_existing_index(index_name, embeddings)

Of course, change the variables “index_name” and “embeddings” to what you named then. Hope this helps!