Datasets:

Modalities:
Text
Formats:
csv
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
License:
stepchoi's picture
Update README.md
531a8bb verified
|
raw
history blame
436 Bytes
metadata
license: apache-2.0

Dataset Summary This dataset is composed of 50 qeustions to test for earnings_10k dataset as part of the LLM-ADE framework (https://arxiv.org/abs/2404.13028), specifically designed to test for native processing/training. We incorporate these qeustions into lm-evalaution-harness (https://github.com/EleutherAI/lm-evaluation-harness) to test our LLM-ADE enhanced model to score if the training was successful.