Having issue about Palm API key with course: Building Multimodal Search and RAG, Document is not helpful

Short Course Q&A Building Multimodal Search and RAGs
1

In the course it keeps using ‘EMBEDDING_API_KEY’ in. the header as below, but the question is how can I get this so called EMBEDDING_API_KEY? Seems the Palm will be deprecated soon and no matter what I tried with Gemini API key or other Google AI keys, it never worked. Any advise? Thanks.

client = weaviate.connect_to_embedded(
version=“1.24.4”,
environment_variables={
“ENABLE_MODULES”: “backup-filesystem,multi2vec-palm”,
“BACKUP_FILESYSTEM_PATH”: “/home/jovyan/work/backups”,
},
headers={
“X-PALM-Api-Key”: EMBEDDING_API_KEY,
}
)

1 Like
2

I’m having a similar problem. Searching for the answer now. If I find something useful, I’ll let you know.

3

Well, I haven’t gotten it working yet, but this Weaviate documentation seems to indicate that version 1.24.4 is not the most recent. The class used that older version which used different headers:

The Weaviate documentation linked above seems to list the steps to get the Vertex AI API key. I’m working through that now.

4

I’m having the same problem here

5

I have documented the steps required for use Vertex AI API key to execute the notebook on Multimodal Search

client = weaviate.connect_to_embedded(
    version="1.24.21",
    environment_variables={
        "ENABLE_MODULES": "backup-filesystem,multi2vec-palm",
        "BACKUP_FILESYSTEM_PATH": "<backup path>"
    },
    headers={
        "X-PALM-Api-Key": EMBEDDING_API_KEY
    }
)