nielsr's picture
nielsr HF Staff
Add link to Github repository, improve description
73f9dc4 verified
|
raw
history blame
427 Bytes
metadata
license: mit

This repository contains a fastText pretraining data filter targeting the LAMBADA task, as discussed in the paper Improving Pretraining Data Using Perplexity Correlations. This filter selects high-quality pretraining data based on correlations between LLM perplexity and downstream benchmark performance.

Code: https://github.com/TristanThrush/perplexity-correlations