Show HN: Semble – Code search for agents that uses 98% fewer tokens than grepgithub.com·8 pts·stephantul·0
Show HN: Semble – Fast code search for agents with near-transformer accuracygithub.com·7 pts·stephantul·0
Show HN: PyNIFE. 400-900× speedup for embedding-based retrieval pipelinesgithub.com·2 pts·stephantul·0
Evaluating ML classifiers using relative error instead of absolute accuracystephantul.github.io·1 pts·stephantul·0
Show HN: SemHash – Fast Semantic Text Deduplication for Cleaner Datasetsgithub.com·6 pts·stephantul·0
Model2Vec: Make sentence transformers 500x faster on CPU, 15x smallerhuggingface.co·5 pts·stephantul·0
Show HN: Model2Vec: make sentence transformers 500x faster on CPU, 15x smallergithub.com·9 pts·stephantul·2
Show HN: Model2Vec: make sentence transformers 500x faster on CPU, 15x smallergithub.com·6 pts·stephantul·2