Add model card
#1
by nielsr HF Staff - opened
README.md
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: transformers
|
| 4 |
+
pipeline_tag: feature-extraction
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
# EvoLen: Evolution-Guided Tokenization for DNA Language Model
|
| 8 |
+
|
| 9 |
+
This repository contains the EvoLen pretrained model described in [EvoLen: Evolution-Guided Tokenization for DNA Language Model](https://huggingface.co/papers/2604.08698).
|
| 10 |
+
|
| 11 |
+
EvoLen is a DNA tokenizer that combines evolutionary stratification with length-aware decoding to better preserve functional sequence units. This model is a BERT-base architecture trained with the EvoLen tokenizer.
|
| 12 |
+
|
| 13 |
+
For more details on the tokenizer, pretraining and fine-tuning, see the [Github repository](https://github.com/HN020719/EvoLen).
|
| 14 |
+
|
| 15 |
+
The checkpoint corresponds to the EvoLen 100k-step pretrained model. Alternative checkpoints and tokenizer files are available on the Hugging Face Hub under the [EvoLenTokenizer organization](https://huggingface.co/EvoLenTokenizer).
|