tsilva/

smoltalk_style_mix-v3

	SmolTalk Style Mix v3




	
	Overview

This dataset mix was produced by datamixxer build. It contains deterministic subsamples from the configured Hugging Face dataset splits. When both train and test files are present, each source bucket was split independently so both output splits preserve the configured blend.

	Versioning and Naming




	

Field Value

Artifact type dataset_mix

Artifact id smoltalk_style_mix

Artifact version v3

Hub repo… See the full description on the dataset page: https://huggingface.co/datasets/tsilva/smoltalk_style_mix-v3.