Getting TypeError when accessing language/use-cases/document-qa/question_answering_documentai_vector_store_palm.ipynb

[SampathkumarSubramaniam]

Hello,
Getting below error when accessing https://github.com/GoogleCloudPlatform/generative-ai/blob/main/language/use-cases/document-qa/question_answering_documentai_vector_store_palm.ipynb

Exact code:

@retry(wait=wait_random_exponential(min=10, max=120), stop=stop_after_attempt(5))
def embedding_model_with_backoff(text=[]):
    **embeddings = embedding_model.get_embeddings(text)**
    return [each.values for each in embeddings][0]

Error log:

 File /lib/python3.11/site-packages/vertexai/language_models/_language_models.py", line 1724, in _prepare_text_embedding_request
    raise TypeError(f"Unsupported text embedding input type: {text}.")
TypeError: Unsupported text embedding input type: nan.

The above exception was the direct cause of the following exception:

Traceback (most recent call last):

    pdf_data_sample["embedding"] =["chunks"].apply(
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Could you help to fix?

[holtskinner]

@SampathkumarSubramaniam Can you provide the input document that caused the issue? (With any PII redacted)
This issue seems to be because the text sent to the embeddings model was empty

[jwchoiKR]

pdf_data_sample = pdf_data_sample.dropna(subset=['chunks'])