Datasets:
Total Token Count & Total Size (GB/TB) in Full Files of bigcode/the-stack-v2-train-smol-ids Dataset
#9
by
Zieksy
- opened
I need to determine the total number of tokens & total size (GB/TB) across all actual files contained within this dataset. Please note, I am referring to the complete files rather than huggingface parquet files such as "train-00000-of-00064.parquet".
Could anyone estimate or find information regarding the total token count for the entire dataset?
Thank you in advance for your help.
Zieksy
changed discussion title from
Total Token Count in Full Files of bigcode/the-stack-v2-train-smol-ids Dataset
to Total Token Count & Total Size (GB/TB) in Full Files of bigcode/the-stack-v2-train-smol-ids Dataset