Hi, I need help with something. I'm trying to load...
# getting-started
d
Hi, I need help with something. I'm trying to load data using tap-csv extractor and then use a mapper using map-gpt-embeddings with the following configuration:
Copy code
mappers:
  - name: map-gpt-embeddings
    variant: meltanolabs
    pip_url: git+<https://github.com/meltanolabs/map-gpt-embeddings>
    executable: map-gpt-embeddings
    settings:
    - name: openai_api_key
      kind: string
      sensitive: true
    mappings:
    - name: add-embeddings
      config:
        document_text_property: name
        document_metadata_property: metadata
        splitter_config:
          chunk_size: 500
          chunk_overlap: 200
I'm just having some trouble configuring the Mapper with
document_metadata_property
. I'm still quite new to all of this; any help would be appreciated.
e
Is
metadata
the right field name coming from tap-csv?
d
How can I know which is
document_metadata_property
in my document? I am still not clear about this configuration.
e
I'm not that too familiar with the assumptions made by the mapper, but it seems like the defaults are set to work with tap-beautifulsoup, which emits records with
page_content
and
metadata
fields: https://github.com/MeltanoLabs/tap-beautifulsoup/blob/5c494852ea64cdcbdf82356e81efbff8b44bc9ce/tap_beautifulsoup/client.py#L91-L95 I guess the mapper then expects your tap to have a similar pair of fields, whose names you can configure with
document_text_property
and
document_metadata_property
respectively.