alt.hn

2/6/2026 at 4:50:11 PM

SMLL: Using 200MB of Neural Network to Save 400 Bytes

https://www.frankchiarulli.com/blog/smll/

by fcjr

2/7/2026 at 4:05:18 AM

No mention of decompression speed and validation, or did I miss something?

by msephton

2/7/2026 at 4:11:40 AM

It's in the post: Benchmarks -> Speed

tl;dr: SMLL is approximately 10,000x slower than Gzip

by savalione