Edit model card

DNA Pre-trained BERT with WordPiece

This is an ineffective model for DNA sequence prediction. The model was pre-trained using reference genomes from multiple plant species with a WordPiece tokenizer. However, the WordPiece tokenizer is not suitable for processing DNA sequences.

This model lacks predictive potential, so this repository will be removed.

Downloads last month
4
Safetensors
Model size
56.2M params
Tensor type
F32
·

Dataset used to train suke-sho/BERT-plant-genome-6

Collection including suke-sho/BERT-plant-genome-6