Instructions to use NextGLab/ORANSight_Qwen_32B_Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NextGLab/ORANSight_Qwen_32B_Instruct with Transformers:

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

pipe = pipeline("text-generation", model="NextGLab/ORANSight_Qwen_32B_Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("NextGLab/ORANSight_Qwen_32B_Instruct")
model = AutoModelForCausalLM.from_pretrained("NextGLab/ORANSight_Qwen_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

vLLM

How to use NextGLab/ORANSight_Qwen_32B_Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "NextGLab/ORANSight_Qwen_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": "NextGLab/ORANSight_Qwen_32B_Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/NextGLab/ORANSight_Qwen_32B_Instruct

SGLang

How to use NextGLab/ORANSight_Qwen_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 "NextGLab/ORANSight_Qwen_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": "NextGLab/ORANSight_Qwen_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 "NextGLab/ORANSight_Qwen_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": "NextGLab/ORANSight_Qwen_32B_Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use NextGLab/ORANSight_Qwen_32B_Instruct with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for NextGLab/ORANSight_Qwen_32B_Instruct to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for NextGLab/ORANSight_Qwen_32B_Instruct to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for NextGLab/ORANSight_Qwen_32B_Instruct to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="NextGLab/ORANSight_Qwen_32B_Instruct",
    max_seq_length=2048,
)

Docker Model Runner
How to use NextGLab/ORANSight_Qwen_32B_Instruct with Docker Model Runner:
```
docker model run hf.co/NextGLab/ORANSight_Qwen_32B_Instruct
```

Model Card for ORANSight Qwen-32B

This model belongs to the first release of the ORANSight family of models.

Developed by: NextG lab@ NC State
License: apache-2.0
Long-context Support up to 128K tokens and can generate up to 8K tokens.
Fine Tuning Framework: Unsloth

Generate with Transformers

Below is a quick example of how to use the model with Hugging Face Transformers:

from transformers import pipeline

# Example query
messages = [
    {"role": "system", "content": "You are an O-RAN expert assistant."},
    {"role": "user", "content": "Explain the E2 interface."},
]

# Load the model
chatbot = pipeline("text-generation", model="NextGLab/ORANSight_Qwen_32B_Instruct")
result = chatbot(messages)
print(result)

Coming Soon

A detailed paper documenting the experiments and results achieved with this model will be available soon. Meanwhile, if you try this model, please cite the below mentioned paper to acknowledge the foundational work that enabled this fine-tuning.

@article{gajjar2024oran,
  title={Oran-bench-13k: An open source benchmark for assessing llms in open radio access networks},
  author={Gajjar, Pranshav and Shah, Vijay K},
  journal={arXiv preprint arXiv:2407.06245},
  year={2024}
}