|
|
||
|---|---|---|
| .. | ||
| Readme.md | ||
| extract_model_cards.py | ||
| filterInvalidTags.py | ||
| hugging2sql.py | ||
| huggingface2.db | ||
| metric_mapping.json | ||
| modality_mapping.json | ||
| model_card_tag_extractor.py | ||
Readme.md
Hugging2KG
Pipeline
- Run
hugging2sql.pyto crawl the Huggingface from their API and save it to SQLite database - Run
tagExtractor.pyto preprocess the stringified list of tags to useful format + filter unimportant (underrepresent tags) - Run
sql2graph.pyto generate the final KG. It is saved inGraphs/graph.ttl