AttributeError: 'ClassTfidfTransformer' object has no attribute 'seed_words'

[archanatikayatray19]

Here's a simple piece of code I am trying to run:

from bertopic import BERTopic


# Train our model with our previously defined models
topic_model = BERTopic(
    embedding_model=embedding_model,
    umap_model=umap_model,
    hdbscan_model=hdbscan_model,
    verbose=True
).fit(abstracts, embeddings)

I get the following error, am I missing something:

2024-11-29 15:24:06,137 - BERTopic - Dimensionality - Fitting the dimensionality reduction algorithm
2024-11-29 15:24:25,821 - BERTopic - Dimensionality - Completed ✓
2024-11-29 15:24:25,823 - BERTopic - Cluster - Start clustering the reduced embeddings
2024-11-29 15:24:26,466 - BERTopic - Cluster - Completed ✓
2024-11-29 15:24:26,466 - BERTopic - Representation - Extracting topics from clusters using representation models.
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-35-3f893b1b089d> in <module>
     10     hdbscan_model=hdbscan_model,
     11     verbose=True
---> 12 ).fit(abstracts, embeddings)

AttributeError: 'ClassTfidfTransformer' object has no attribute 'seed_words'

I am using pip install bertopic>=0.16.0 to install the library.

Thanks! :)

[MaartenGr]

Thanks for sharing! You might have missed this, but you could also open up an issue as that seems to relate a bit more to your question.

That said, there are a couple of things missing. Could you share the exact version of BERTopic? Also, could you share the full error log?