---
license: mit
language:
- en
---

# DNA → Amino Acid Translation Dataset

### Description
This dataset contains randomly generated DNA sequences (A, C, G, T)
and their amino-acid translations using the standard genetic code.
It is suitable for Seq2Seq learning and translation model training.

All sequence lengths are multiples of 3 to valid codon structure and vary widely to ensure diversity.
- **Minimum length:** 60 nucleotides  
- **Maximum length:** 1024 nucleotides 

### Dataset Structure
- **Fields**
  - `src`: input DNA sequence (string)
  - `tgt`: translated amino acid sequence (string)
- **Splits**
  - `train`, `validation`, `test`