weaviate.exceptions.UnexpectedStatusCodeException: batch response! Unexpected status code: 400, with response body: {'code': 400, 'message': 'parsing body body from "" failed, because json: cannot unmarshal array into Go struct field Object.objects.vector

Support
1

from gensim.models.doc2vec import Doc2Vec, TaggedDocument
from nltk.tokenize import word_tokenize

def embedder(chunks):
# Preprocess the documents and create TaggedDocuments
tagged_data = [TaggedDocument(words=word_tokenize(chunk.lower()), tags=[str(i)]) for i, chunk in enumerate(chunks)]

# Train the Doc2Vec model
model = Doc2Vec(vector_size=100, min_count=1)
model.build_vocab(tagged_data)
model.train(tagged_data, total_examples=model.corpus_count, epochs=model.epochs)

# Verify the vector size
print("Vector size:", model.vector_size)

embeddings = []

for i, chunk in enumerate(chunks):
    # Infer vector for the entire chunk
    chunk_vector = model.infer_vector(word_tokenize(chunk.lower()))
    embeddings.append(chunk_vector)

return embeddings

This is my embedding that goes in a chunking code. When i saved my chunk and vector in a json file i ask. I wanted to save it in a weavite database.

import weaviate
import json
import pprint
f = open(‘data.json’)
data = json.load(f)

client = weaviate.Client(
url = “http://10.0.129.19:8888”, # Replace with your endpoint
)

Configure a batch process

client.batch.configure(batch_size=100) # Configure batch
with client.batch as batch:
# Batch import all Questions
for i, d in enumerate(data):
print(f"importing answers: {i+1}")

    # pprint.pprint(d["Answer"])
    properties = {
        "answer": d["Answer"]
    }

    batch.add_data_object(properties, "QWERTY", vector=d["Vector"])

in this code what worked with a other embedding. Is it my embedding code?

2

Hi!

This indeed seems like a bad vector ingestion.

What is the version you are running?

can you make sure that d[“Vector”] has a valid vector?

also, can you manually replace it with a known vector?

Thanks!