By accessing this dataset, you agree to use it responsibly and ethically. You agree not to use the dataset for any form of bioterrorism, to harm bees or other pollinators, to disrupt ecosystems, or to commit any act that negatively impacts biodiversity or public health. You also agree not to use this dataset to develop technologies or conduct experiments that could cause harm to humans, animals, or the environment.

Dataset Card for "bees-internal"

Full length OCR of Bee material and other Lore. Documents are split into multiple chunks if they contain more than 0.5 MB of text, to avoid destroying the CPU during tokenization.

counts

all counts are for the default config.

tokens

GPT-4 tiktoken token count:

         token_count
count    1384.000000
mean    32776.423410
std     33652.185553
min       215.000000
25%      3514.000000
50%      8942.000000
75%     65717.750000
max    160077.000000

Total count: 45.36 M tokens

meta-llama/Meta-Llama-3-8B token count:

         token_count
count    1384.000000
mean    31807.650289
std     32651.250770
min       212.000000
25%      3412.750000
50%      8733.500000
75%     64047.000000
max    156381.000000

Total count: 44.02 M tokens

input files

INFO: Found 1457 text files - 2024-Feb-20_13-19
INFO: Train size: 1384
Validation size: 36
Test size: 37

Downloads last month: 36

Models trained or fine-tuned on BEE-spoke-data/bees-internal