Skip to content

unstructured-ingest weaviate example is broken #269

Open
@hsm207

Description

@hsm207

From the docs, this command should work:

#!/usr/bin/env bash

# Chunking and embedding is optional.

unstructured-ingest \
  local \
    --input-path $LOCAL_FILE_INPUT_DIR \
    --output-dir $LOCAL_FILE_OUTPUT_DIR \
    --strategy hi_res \
    --chunk-elements \
    --embedding-provider huggingface \
    --num-processes 2 \
    --verbose \
    --strategy fast \
    --partition-by-api \
    --api-key $UNSTRUCTURED_API_KEY \
    --partition-endpoint $UNSTRUCTURED_API_URL \
    --additional-partition-args="{\"split_pdf_page\":\"true\", \"split_pdf_allow_failed\":\"true\", \"split_pdf_concurrency_level\": 15}" \
  weaviate \
    --host-url $WEAVIATE_URL \
    --api-key $WEAVIATE_API_KEY \
    --class-name $WEAVIATE_COLLECTION_CLASS_NAME

but it throws:

Error: No such option: --chunk-elements (Possible options: --chunk-include-orig-elements, --chunk-overlap, --no-chunk-include-orig-elements)

using the following libraries:

Name: unstructured
Version: 0.16.6

Name: unstructured-ingest
Version: 0.3.4

Name: weaviate-client
Version: 4.9.4

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions