Instructions to use g-h-chen/nv_rsi with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use g-h-chen/nv_rsi with Transformers:

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

pipe = pipeline("image-text-to-text", model="g-h-chen/nv_rsi")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("g-h-chen/nv_rsi")
model = AutoModelForMultimodalLM.from_pretrained("g-h-chen/nv_rsi", device_map="auto")
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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use g-h-chen/nv_rsi with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "g-h-chen/nv_rsi"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "g-h-chen/nv_rsi",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/g-h-chen/nv_rsi

SGLang

How to use g-h-chen/nv_rsi 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 "g-h-chen/nv_rsi" \
    --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": "g-h-chen/nv_rsi",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "g-h-chen/nv_rsi" \
        --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": "g-h-chen/nv_rsi",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use g-h-chen/nv_rsi with Docker Model Runner:
```
docker model run hf.co/g-h-chen/nv_rsi
```

Nova RSI MU-182 corrected checkpoint

This is the corrected completion-only, one-exposure MU-182 checkpoint trained at learning rate 1e-5. The archival original all-turn checkpoint is available on the mu182-allturn-3ep revision of this repository.

On Unified256, the T3 baseline scored 48/256 functionality and 5/256 attacks; this corrected checkpoint scored 53/256 functionality and 5/256 attacks. Against T3 CA, attacks fell from 3 to 1. The checkpoint therefore recovered functionality but did not improve the intended security behavior.

This revision contains 15 loadable safetensor shards plus configuration and tokenizer metadata. Optimizer state and duplicate trainer exports are excluded.

Data and evidence: https://huggingface.co/datasets/g-h-chen/nv_rsi
Code and notes (private): https://github.com/Cyphysecurity/rsi_nova

Downloads last month: 17

Safetensors

Model size

35B params

Tensor type

BF16