Hugging Face's logo

Datasets:
Compactbot
/
slm-architecture-benchmark-specs

Modalities:
Tabular
Text
Formats:
json
Languages:
English
Size:
< 1K
Tags:
small-language-model
slm
benchmark
evaluation
methodology
perplexity
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
slm-architecture-benchmark-specs / examples.jsonl

Commit History

Add worked examples: real numbers from verified benchmarks (byte-PPL, tied-dedup, held-out PPL)
7753e71
verified

Compactbot commited on