README.md

metadata

license: mit

FLAN v2 Cot Deduplicated Dataset

Data Preprocessing

• Remove instructions with less than 100 in 'targets'.
• Dedepulicate Dataset using cosine similarity with a threshold of 0.95.

Code

• Github repo : https://github.com/AJlearner46/Deduplicate-flanv2-finetune-LLaMa3-

Acknowledgments

• The original dataset is provided by SirNeural/flan_v2.
• Tokenizer used: bert-base-uncased from Hugging Face.