Evo 2 1B base DNA language model weights (8K context)

Provider
Arc Institute
License
Apache-2.0 (source)
Size
4 files
Format
JSON (in sampled files)
Stored at
Hugging Face (https://huggingface.co/arcinstitute/evo2_1b_base)
Access
Free and public. No account with the data provider is needed.
Category
Genomics & Bioinformatics
Source
https://www.biorxiv.org/content/10.1101/2025.02.18.638918v1

About Evo 2 1B base DNA language model weights (8K context)

Open weights for Evo 2 1B base, the smallest checkpoint of Arc Institute's Evo 2 DNA language model family: 25 layers and 1 billion parameters, trained on OpenGenome2 with an 8,192 base pair context. It is the practical size for trying variant-effect scoring and sequence embeddings on a single GPU before moving to the 7B or 40B checkpoints. The mounted repo holds evo2_1b_base.pt (2.7 GB) and config.json, loaded with the evo2 package from the mount path. The OpenGenome2 training corpus is also in this catalog. Apache 2.0.

How to load Evo 2 1B base DNA language model weights (8K context) in Python

import os, json
base = "/mnt/archive"   # this model repo, mounted read-only
files = sorted(os.listdir(base)); print(files)
print(json.dumps(json.load(open(os.path.join(base, "config.json"))), indent=1)[:800])
print("evo2_1b_base.pt", os.path.getsize(os.path.join(base, "evo2_1b_base.pt")) // 2**20, "MB")

Written for a Scigantic notebook, where /mnt/archive is the mounted dataset.

Data structure

File types in sampled files: JSON

Files

Example file: evo2_1b_base.pt

Analyze this dataset without downloading it

On Scigantic, Evo 2 1B base DNA language model weights (8K context) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.

This page is also available as Markdown and as JSON.

Related Genomics & Bioinformatics datasets