Instructions to use OrionLLM/Nebula with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OrionLLM/Nebula with Transformers:

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

pipe = pipeline("text-generation", model="OrionLLM/Nebula")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OrionLLM/Nebula")
model = AutoModelForCausalLM.from_pretrained("OrionLLM/Nebula")
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 OrionLLM/Nebula with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/OrionLLM/Nebula

SGLang

How to use OrionLLM/Nebula 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 "OrionLLM/Nebula" \
    --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": "OrionLLM/Nebula",
		"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 "OrionLLM/Nebula" \
        --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": "OrionLLM/Nebula",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use OrionLLM/Nebula with Docker Model Runner:
```
docker model run hf.co/OrionLLM/Nebula
```

DedeProGames commited on Apr 6

Commit

40a5695

verified ·

1 Parent(s): eaf1738

Update README.md

Browse files

Files changed (1) hide show

README.md +2 -3

README.md CHANGED Viewed

@@ -20,7 +20,7 @@ tags:
 </p>
 ## 1. Introduction
-**Nebula** is a **321M-parameter** generalist Small Reasoning Model trained on **200B+ tokens**.
 Nebula is designed to deliver an unusually strong balance of **memory**, **general reasoning**, **math**, and **retrieval-friendly behavior** for its size class, aiming to outperform many small models of a similar parameter range on non-code, industry-style benchmarks.
@@ -61,8 +61,7 @@ Traces use the following stenographic notation integrated into special tokens:
 This reasoning format is designed to remain expressive while being lightweight enough for a small model.
 ## 3. Fine-Tuning/RL
-Nebula has been successfully fine-tuned for a variety of tasks including text classification and
-<a href="https://x.com/darrenangle/status/1990259914602856831">poetry writing</a>.
 Because Nebula is a reasoning-oriented model, it is expected to train well with reinforcement learning methods such as **GRPO**, both for **verifiable tasks** (with objective rewards) and for subjective tasks using an **LLM-as-a-judge**.

 </p>
 ## 1. Introduction
+**Nebula** is a **320M-parameter** generalist Small Reasoning Model trained on **200B+ tokens**.
 Nebula is designed to deliver an unusually strong balance of **memory**, **general reasoning**, **math**, and **retrieval-friendly behavior** for its size class, aiming to outperform many small models of a similar parameter range on non-code, industry-style benchmarks.
 This reasoning format is designed to remain expressive while being lightweight enough for a small model.
 ## 3. Fine-Tuning/RL
+Nebula has been successfully fine-tuned for a variety of tasks
 Because Nebula is a reasoning-oriented model, it is expected to train well with reinforcement learning methods such as **GRPO**, both for **verifiable tasks** (with objective rewards) and for subjective tasks using an **LLM-as-a-judge**.