Datasets:

crumb
/

c4-benchfilter-nano

Tasks:

Text Generation

Modalities:

Formats:

Sub-tasks:

language-modeling

masked-language-modeling

Languages:

Size:

Tags:

Libraries:

License:

Dataset card Viewer Files Files and versions Community

crumb commited on Oct 22, 2023

Commit

8a0aa25

•

1 Parent(s): a811938

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -35,7 +35,7 @@ configs:
 A derivation of the first 3M samples from the C4 dataset for a cheap continued pretraining addition.
-The estimated top 10% of highest n-token (mean 3,4,5) overlaps for each of the
 selected benchmark datasets (arc, truthful_qa, hellaswag, mmlu, humaneval) based
 on 1k samples, within the first 3M samples of C4. The top scoring sample
 datasets for each benchmark are then filtered again for top 30% scores and

 A derivation of the first 3M samples from the C4 dataset for a cheap continued pretraining addition.
+The estimated top 10% of highest estimated length normalized ngram (mean of tri, quad, and penta-gram) overlaps for each of the
 selected benchmark datasets (arc, truthful_qa, hellaswag, mmlu, humaneval) based
 on 1k samples, within the first 3M samples of C4. The top scoring sample
 datasets for each benchmark are then filtered again for top 30% scores and