seanpedrickcase/topic_modelling at e1c1f680dea034147945c492dab6d68c4d1cb489

topic_modelling / funcs

116 kB

4 contributors

Reduce outliers now more efficient and relabels with correct vectoriser. Default topic labels now tidier. Hiearchical topics outputs more useful for joining to df afterwards. Switched low resource reduction algorithm to UMAP as default is not good.

e1c1f68 almost 2 years ago

__init__.py
0 Bytes

first commit almost 2 years ago
anonymiser.py
10.2 kB

Added clean data options, improved re-representation options and visualisation. General format changes almost 2 years ago
bertopic_vis_documents.py
47.2 kB

Reduce outliers now more efficient and relabels with correct vectoriser. Default topic labels now tidier. Hiearchical topics outputs more useful for joining to df afterwards. Switched low resource reduction algorithm to UMAP as default is not good. almost 2 years ago
clean_funcs.py
5.03 kB

Reduce outliers now more efficient and relabels with correct vectoriser. Default topic labels now tidier. Hiearchical topics outputs more useful for joining to df afterwards. Switched low resource reduction algorithm to UMAP as default is not good. almost 2 years ago
embeddings.py
2.54 kB

Hopefully now LLM download from hub should work almost 2 years ago
helper_functions.py
9.91 kB

Should now parse custom regex correctly. Will now wipe previously created embeddings if 'low resource mode' option switched. almost 2 years ago
presidio_analyzer_custom.py
4.18 kB

Added clean data options, improved re-representation options and visualisation. General format changes almost 2 years ago
prompts.py
4.86 kB

Model export changed to safetensors. Improved representational model function. Got zero shot topic modelling working almost 2 years ago
representation_model.py
6.74 kB

Hopefully now LLM download from hub should work almost 2 years ago
topic_core_funcs.py
25.1 kB

Reduce outliers now more efficient and relabels with correct vectoriser. Default topic labels now tidier. Hiearchical topics outputs more useful for joining to df afterwards. Switched low resource reduction algorithm to UMAP as default is not good. almost 2 years ago