Instructions to use marin-dna/marin-dna-scaling-v0.5-h896-p128M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use marin-dna/marin-dna-scaling-v0.5-h896-p128M with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="marin-dna/marin-dna-scaling-v0.5-h896-p128M")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("marin-dna/marin-dna-scaling-v0.5-h896-p128M")
model = AutoModelForCausalLM.from_pretrained("marin-dna/marin-dna-scaling-v0.5-h896-p128M", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use marin-dna/marin-dna-scaling-v0.5-h896-p128M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "marin-dna/marin-dna-scaling-v0.5-h896-p128M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "marin-dna/marin-dna-scaling-v0.5-h896-p128M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/marin-dna/marin-dna-scaling-v0.5-h896-p128M

SGLang

How to use marin-dna/marin-dna-scaling-v0.5-h896-p128M with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "marin-dna/marin-dna-scaling-v0.5-h896-p128M" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "marin-dna/marin-dna-scaling-v0.5-h896-p128M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "marin-dna/marin-dna-scaling-v0.5-h896-p128M" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "marin-dna/marin-dna-scaling-v0.5-h896-p128M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use marin-dna/marin-dna-scaling-v0.5-h896-p128M with Docker Model Runner:
```
docker model run hf.co/marin-dna/marin-dna-scaling-v0.5-h896-p128M
```

MarinDNA v0.5 scaling ladder — 128M

This 128,484,480-parameter nucleotide-level causal language model is a member of the eight-model MarinDNA v0.5 parameter-scaling ladder developed with Marin. This repository contains only the final step-215573 checkpoint from run dna-bolinas-scaling-v0.5-h896-p128M-43ec40, with its tokenizer bundled. It accompanies A 1B standard Transformer rivals Evo 2 40B on variant effect prediction.

Model details

Field	Value
Architecture	Qwen3-compatible decoder-only Transformer
Parameters	128,484,480
Layers	10
Hidden / intermediate size	896 / 3,584
Attention heads / KV heads	7 / 7
Context	256 tokens: one BOS token followed by up to 255 DNA bases
Checkpoint	Final checkpoint, step 215573 (after 215,574 zero-indexed training steps)
Token exposure	approximately 84.77B nucleotide tokens
Stored weight dtype	float32
License	Apache-2.0

The canonical source checkpoint is gs://marin-us-east5/checkpoints/dna-bolinas-scaling-v0.5-h896-p128M-43ec40/hf/step-215573; the byte-identical evals_v2 transfer cache is s3://oa-bolinas/snakemake/analysis/evals_v2/results/checkpoints/scaling-v0.5-h896-p128M-step-215573. The commit-pinned training script defines the production ladder.

Loading

from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "marin-dna/marin-dna-scaling-v0.5-h896-p128M"
tokenizer = AutoTokenizer.from_pretrained(repo_id)
model = AutoModelForCausalLM.from_pretrained(repo_id)

Tokenizer and input format

The bundled tokenizer is case-insensitive and represents one nucleotide per token. Its vocabulary is [PAD]=0, [UNK]=1, [BOS]=2, a=3, c=4, g=5, t=6. Pass raw DNA strings containing A, C, G, and T without spaces. The tokenizer lowercases input and prepends [BOS]; it has no EOS token. Other symbols map to [UNK]. Because BOS occupies one of the 256 positions, inputs are limited to 255 DNA bases.

Training data and protocol

Every model in the ladder used the same batch size, token budget, optimizer hyperparameters, tokenizer, and three-way training mixture: 73.19% CDS, 20.62% upstream, and 6.19% downstream sequence. Lowercase soft-masked positions received loss weight 0.01, versus 1.0 for uppercase positions. The ladder changes model scale, not the training-data recipe.

Training datasets: CDS, upstream, and downstream. The matched CDS, upstream, and downstream datasets were validation probes, not training data. The exact tokenizer, data, and mixture definitions are commit-pinned here.

Downloads last month: 70

Safetensors

Model size

0.1B params

Tensor type

F32

Collection including marin-dna/marin-dna-scaling-v0.5-h896-p128M

A 1B standard Transformer rivals Evo 2 40B on VEP

Collection

m5.1 release; scaling ladder; 5 training datasets; 5 validation probes; Mendelian/SGE evaluation only. Blog: https://openathena.ai/blog/marin-dna/ • 21 items • Updated 3 days ago • 2