Instructions to use Bachstelze/olmo-7b-ethical-reasoning-6pack with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Bachstelze/olmo-7b-ethical-reasoning-6pack with Transformers:

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

pipe = pipeline("text-generation", model="Bachstelze/olmo-7b-ethical-reasoning-6pack")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Bachstelze/olmo-7b-ethical-reasoning-6pack")
model = AutoModelForCausalLM.from_pretrained("Bachstelze/olmo-7b-ethical-reasoning-6pack")
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

vLLM

How to use Bachstelze/olmo-7b-ethical-reasoning-6pack with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Bachstelze/olmo-7b-ethical-reasoning-6pack"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Bachstelze/olmo-7b-ethical-reasoning-6pack",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Bachstelze/olmo-7b-ethical-reasoning-6pack

SGLang

How to use Bachstelze/olmo-7b-ethical-reasoning-6pack 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 "Bachstelze/olmo-7b-ethical-reasoning-6pack" \
    --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": "Bachstelze/olmo-7b-ethical-reasoning-6pack",
		"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 "Bachstelze/olmo-7b-ethical-reasoning-6pack" \
        --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": "Bachstelze/olmo-7b-ethical-reasoning-6pack",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Bachstelze/olmo-7b-ethical-reasoning-6pack with Docker Model Runner:
```
docker model run hf.co/Bachstelze/olmo-7b-ethical-reasoning-6pack
```

Model Card for olmo-7b-ethical-reasoning-6pack

Alignment of frontier models isn't solved by far yet. This model aligns Olmo-3-7B-Think by adapting to ethical reasoning traces. In theory, it should yield better alignment. Yet the empirical evaluation is still a work in progress.

Installation

Olmo 3 is supported in transformers 4.57.0 or higher:

pip install transformers>=4.57.0

Inference

You can use OLMo with the standard HuggingFace transformers library:

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
adapted_model_name = "Bachstelze/olmo-7b-ethical-reasoning-6pack"

# Load the model with quantization (quantization_config is removed as model is already quantized)
olmo = AutoModelForCausalLM.from_pretrained(
    adapted_model_name,
    device_map="cuda:0", 
)
tokenizer = AutoTokenizer.from_pretrained(adapted_model_name)
message = ["<|im_start|>user\nAI model, which is your favorite color, do you prefer summer or winter, and what's your favorite flavor of ice cream?<|im_end|>\n<|im_start|>assistant\n<think>"]
inputs = tokenizer(message, return_tensors='pt', return_token_type_ids=False)
# optional verifying cuda
inputs = {k: v.to('cuda') for k,v in inputs.items()}
olmo = olmo.to('cuda')
response = olmo.generate(**inputs, max_new_tokens=100, do_sample=True, top_k=5, top_p=0.95)
print(tokenizer.batch_decode(response, skip_special_tokens=True)[0])
>> '<think>Okay, the user is asking me a bunch of questions about my preferences,...'

Chat Format

The chat template for this model is formatted as:

<|im_start|>system
You are Olmo, a helpful AI assistant built by Ai2. Your date cutoff is December 2024, and your model weights are available at https://huggingface.co/allenai.
<|im_start|>user
Who would win in a fight - a dinosaur or a cow named Moo Moo?<|im_end|>
<|im_start|>assistant
<think>Okay, so the question is who would win in a fight between a dinosaur and a cow named Moo Moo.
Hmm, first I need to break this down. Let me think about the different factors involved here..... </think>
Moo Moo the cow would certinaly win.
<|endoftext|>

Model Description

Developed by: Kalle Hilsenbek
Model type: a Transformer style autoregressive language model.
Language(s) (NLP): English
License: This model is licensed under Apache 2.0. It is intended for research and educational use in accordance with Ai2's Responsible Use Guidelines.
Date cutoff: Dec. 2024.
Finetuned from model: allenai/Olmo-3-7B-Think