Instructions to use allenai/OLMo-2-0325-32B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use allenai/OLMo-2-0325-32B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="allenai/OLMo-2-0325-32B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("allenai/OLMo-2-0325-32B-Instruct")
model = AutoModelForCausalLM.from_pretrained("allenai/OLMo-2-0325-32B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use allenai/OLMo-2-0325-32B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "allenai/OLMo-2-0325-32B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "allenai/OLMo-2-0325-32B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/allenai/OLMo-2-0325-32B-Instruct

SGLang

How to use allenai/OLMo-2-0325-32B-Instruct 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 "allenai/OLMo-2-0325-32B-Instruct" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "allenai/OLMo-2-0325-32B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "allenai/OLMo-2-0325-32B-Instruct" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "allenai/OLMo-2-0325-32B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use allenai/OLMo-2-0325-32B-Instruct with Docker Model Runner:
```
docker model run hf.co/allenai/OLMo-2-0325-32B-Instruct
```

32b, 4k ctx?

by lucyknada - opened Mar 13, 2025

Discussion

lucyknada

Mar 13, 2025

is 4k the final context-length planned for this model? or is there more in the works?

Araki

Mar 13, 2025

I really like what they did with the the whole "fully open source" deal, but the 4k context length is indeed head-scratching. I'd also like to hear a word on this.

lucyknada

Mar 13, 2025

self replying, since apparently it was mentioned on twitter but not here, what "very soon" means is another question.

MaziyarPanahi

Mar 15, 2025

I can’t serve this model with a context length limited to 4K. A 4K context might be acceptable for smaller models (0.5B or 1B) intended for on-device use cases, but for a 32B model, I need it to support at least a 128K context window to achieve decent performance at 32K.

deleted

Mar 19, 2025

Anyone else getting

ollama run MHKetbi/allenai_OLMo2-0325-32B-Instruct:Q8_0 Error: llama runner process has terminated: error loading model: check_tensor_dims: tensor 'blk.0.attn_k_norm.weight' has wrong shape; expected 5120, got 1024, 1, 1, 1

amanrangapur

Mar 20, 2025

•

edited Mar 20, 2025

Hi @lucyknada , we’re working on making the context longer. We’re definitely planning to do that in the next versions. Stay tuned! Thanks everyone else for the feedback.

amanrangapur

Mar 20, 2025

Anyone else getting

ollama run MHKetbi/allenai_OLMo2-0325-32B-Instruct:Q8_0 Error: llama runner process has terminated: error loading model: check_tensor_dims: tensor 'blk.0.attn_k_norm.weight' has wrong shape; expected 5120, got 1024, 1, 1, 1

Hey @TheSeminal, there is this issue on going with llama.cpp. Check this out for more context: https://huggingface.co/allenai/OLMo-2-0325-32B-Instruct-GGUF/discussions/1

mruderman

May 14, 2025

Hi @lucyknada , we’re working on making the context longer. We’re definitely planning to do that in the next versions. Stay tuned! Thanks everyone else for the feedback.

That's fantastic to hear. Any chance there is a rough timeline for when that might happen? Thank you again for all you do!

amanrangapur

May 16, 2025

Hey @mruderman , not sure when we're going to release it. But we're grinding on it.

baileyk

Ai2 org Jul 17, 2025

Hi! Thanks again for the inquiry. We’re currently working on closing out old tickets, so we’re closing this out for now, but if you’d still like an answer, please re-open and we will get back to you!

baileyk changed discussion status to closed Jul 17, 2025

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment