BenchBloom
BenchBloom - a forensic herbarium for benchmark contamination: exact n-gram/Jaccard and containment scans in pure Python, plus a C++17 MinHash engine for finding benchmark items leaked into training corpora.
- benchmark
- cli
- containment
- contamination
- cpp
- data-science
- deduplication
- evaluation
- jaccard
- leakage
- machine-learning
- minhash
- ngram
- python
- text-processing
- Stars
- 105
- Forks
- 33
- + today
- +2
- Created
- 9y
Ranking data as of October 11, 2026 (UTC).
Star History
Today, hour by hour
02
Overview
BenchBloom - a forensic herbarium for benchmark contamination: exact n-gram/Jaccard and containment scans in pure Python, plus a C++17 MinHash engine for finding benchmark items leaked into training corpora. It ranks #404 on GitTiger, gaining +2 stars on October 11, 2026 (UTC).
The project is written in Python and has 33 forks. It was created 9y ago.
Installation
git clone https://github.com/zizimarxgeoleh/BenchBloom.git
cd BenchBloom
# see README for setup