Datasets:
Clusters in different subsets, clusters' ratios and annotations
#3
by
gvlassis
- opened
Hello!
Thanks a ton for detokenizing ClimbMix!
I build on top of your work, and:
- Separated clusters in different subsets (so you do not have to load everything in memory)
- Calculated precise cluster ratios (useful for subsampling)
- Added topic annotations
Here is my repo: https://huggingface.co/datasets/gvlassis/ClimbMix