Train a dictionary for small files

Train a dictionary on similar small files, then compress each file with it. The ratio gets much better

zstd --train samples/*.json -o dict.zstd && zstd -D dict.zstd file.json

tool
zstd2 commands
tags
zstdperformancedisk

More zstd commands

  • zstd -19 --long=27 -T0 huge.bin -o huge.bin.zst

    Level 19 with a 128 MB long-distance window on all cores. Very good ratio for logs and backups

all 2 zstd commands