search

bigcode-project / bigcode-analysis

Repository for analysis and experiments in the BigCode project.
Apache License 2.0
113 stars 20 forks source link

Minhash Improvement #30

Closed ChenghaoMou closed 1 year ago

ChenghaoMou commented 1 year ago
ChenghaoMou commented 1 year ago

This also includes default parameter change from unigram to 5-gram, 0.7 threshold, and stats from our experiments. (#10 #7)