MaartenGr / MaartenGr/BERTopic
Reproducibility issue in Guided Topic Modeling
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 7.8k
- Forks
- 920
- Avg merge
- 22h 24m
- Merged PRs (30d)
- 5
Description
### Have you searched existing issues? 🔎
- [x] #2278
### Desribe the bug
Hi,
I'm running a topic model and passing a seed_topic_list. This helps me generate more meaningful topics, but I cannot reproduce the model.
UMAP's random_state is fixed. This doesn't happen when seed_topic_list is not passed, so I assume it's related to this parameter.
I wonder why this happens and if it's intrinsic to the method. I couldn't find related info, so I'd appreciate some insight into this.
Best,
### BERTopic Version
0.16.4
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the Guided Topic Modeling entry point and reproduce the issue using seed_topic_list with UMAP's fixed random_state, comparing it with a run without seed_topic_list. Confirm whether repeated runs produce different models and document the reproducibility behavior for BERTopic 0.16.4.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100