Instructions to use cpral/Nex-N2-Pro-EXL3-3.2BPW with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cpral/Nex-N2-Pro-EXL3-3.2BPW with Transformers:

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

pipe = pipeline("text-generation", model="cpral/Nex-N2-Pro-EXL3-3.2BPW")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("cpral/Nex-N2-Pro-EXL3-3.2BPW")
model = AutoModelForMultimodalLM.from_pretrained("cpral/Nex-N2-Pro-EXL3-3.2BPW")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use cpral/Nex-N2-Pro-EXL3-3.2BPW with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cpral/Nex-N2-Pro-EXL3-3.2BPW"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cpral/Nex-N2-Pro-EXL3-3.2BPW",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/cpral/Nex-N2-Pro-EXL3-3.2BPW

SGLang

How to use cpral/Nex-N2-Pro-EXL3-3.2BPW 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 "cpral/Nex-N2-Pro-EXL3-3.2BPW" \
    --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": "cpral/Nex-N2-Pro-EXL3-3.2BPW",
		"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 "cpral/Nex-N2-Pro-EXL3-3.2BPW" \
        --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": "cpral/Nex-N2-Pro-EXL3-3.2BPW",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use cpral/Nex-N2-Pro-EXL3-3.2BPW with Docker Model Runner:
```
docker model run hf.co/cpral/Nex-N2-Pro-EXL3-3.2BPW
```

Configuration Parsing Warning:In config.json: "quantization_config.bits" must be an integer

~3.2BPW custom optimized EXL3 quant of Nex-N2-Pro 397B.

 -- A perplexity:  3.27223557
 -- B perplexity:  3.31265510
 -- A label in top-K:
      K = 1: 0.7130
      K = 2: 0.8113
      K = 3: 0.8532
      K = 4: 0.8769
      K = 5: 0.8923
 -- B label in top-K:
      K = 1: 0.7099
      K = 2: 0.8094
      K = 3: 0.8520
      K = 4: 0.8755
      K = 5: 0.8920
 -- Top-K agreement, A vs B:
      K = 1: 0.9426
      K = 2: 0.7747
      K = 3: 0.5770
      K = 4: 0.4003
      K = 5: 0.2656
 -- KL divergence (A, B):  0.04183738
 -- KL divergence (B, A):  0.03957721

Downloads last month: 2

Model tree for cpral/Nex-N2-Pro-EXL3-3.2BPW

Base model

nex-agi/Nex-N2-Pro

Quantized

(23)

this model