Training data

[zbruhnke]

tinystories_data00.bin

tiny_stories_data_pretokenized.zip

stories110M.bin

All seem to be missing here which would be needed to run the benchmarks?

[jamiej]

Guessing this is the training set

• https://www.kaggle.com/datasets/varunsenthilb20ee077/pretokenized-tinystories

Note only the first 40MB is used.