Datasets:

slmconsortium
/

OpenThink

Tasks:

Text Generation

Question Answering

Modalities:

Formats:

Languages:

Size:

Tags:

chain-of-thought

small-language-models

Libraries:

License:

Dataset card Data Studio Files Files and versions

KlondikeDev commited on about 2 hours ago

Commit

6044a12

·

verified ·

1 Parent(s): 57bbeef

Card: restore the V100 in the description

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -90,7 +90,7 @@ Agreement between two solves is a filter, not a proof: some rows will still have
 ## Results
-Fine-tuning on OpenThink measurably helps small models. Fine-tuning the 144M-parameter [Boris-2](https://huggingface.co/opencerebral/Boris-2) on the hybrid config (2 epochs, about 5 minutes) gave [Boris-2-Think](https://huggingface.co/opencerebral/Boris-2-Think):
 | Benchmark | Boris-2 | Boris-2-Think |
 |---|---:|---:|

 ## Results
+Fine-tuning on OpenThink measurably helps small models. Fine-tuning the 144M-parameter [Boris-2](https://huggingface.co/opencerebral/Boris-2) on the hybrid config (2 epochs, about 5 minutes on one V100) gave [Boris-2-Think](https://huggingface.co/opencerebral/Boris-2-Think):
 | Benchmark | Boris-2 | Boris-2-Think |
 |---|---:|---:|