Instructions to use neural-bulos/nebulos-concise with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use neural-bulos/nebulos-concise with Transformers:

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

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

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("neural-bulos/nebulos-concise")
model = AutoModelForCausalLM.from_pretrained("neural-bulos/nebulos-concise")
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 neural-bulos/nebulos-concise with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/neural-bulos/nebulos-concise

SGLang

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

Docker Model Runner
How to use neural-bulos/nebulos-concise with Docker Model Runner:
```
docker model run hf.co/neural-bulos/nebulos-concise
```

Nebulos-Concise

A fine-tuned Qwen3-4B model that gives concise, no-fluff answers for development tasks.

Overview

Nebulos-Concise is a dev assistant that cuts the garbage. No "Sure! Let me help you with that." No unnecessary explanations. No filler. You ask for code, you get code.

Example:

User: Write a Python function to check if a number is even.

Nebulos-Concise: k.
def is_even(n):
    return n % 2 == 0

Training Details

Base model: Qwen/Qwen3-4B
Method: QLoRA fine-tuning + adapter merge
LoRA config: r=32, alpha=64, dropout=0.05
Training data: 32 curated prompt-response pairs
Epochs: 3
Hardware: Google Colab T4 GPU

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("neural-bulos/nebulos-concise")
tokenizer = AutoTokenizer.from_pretrained("neural-bulos/nebulos-concise")

messages = [
    {"role": "system", "content": "You are Nebulos, a concise dev assistant. You give short, direct answers with no filler, no pleasantries, and no unnecessary explanation. When asked for code, respond with minimal text and the code. Never say 'Sure', 'Certainly', 'I'd be happy to', or anything like that. Just the answer."},
    {"role": "user", "content": "Write a function to reverse a list in Python."},
]

inputs = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt", enable_thinking=False)
input_ids = inputs["input_ids"].to(model.device)
outputs = model.generate(input_ids, max_new_tokens=256, temperature=0.7, do_sample=True, pad_token_id=tokenizer.eos_token_id)
print(tokenizer.decode(outputs[0][input_ids.shape[1]:], skip_special_tokens=True))

GGUF

A GGUF quantized version is available in this repository as model-q4_k_m.gguf.

License

Apache 2.0 (inherited from Qwen3-4B)

Credits

Fine-tuned by neural-bulos using QLoRA on Colab T4.

Downloads last month: 36

Safetensors

Model size

4B params

Tensor type

F32

F16

Model tree for neural-bulos/nebulos-concise

Base model

Qwen/Qwen3-4B-Base

Finetuned

Qwen/Qwen3-4B

Quantized

(240)

this model