Instructions to use llm-jp/llm-jp-4-vl-9b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use llm-jp/llm-jp-4-vl-9b with Transformers:

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

pipe = pipeline("image-text-to-text", model="llm-jp/llm-jp-4-vl-9b", trust_remote_code=True)
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 AutoModel
model = AutoModel.from_pretrained("llm-jp/llm-jp-4-vl-9b", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use llm-jp/llm-jp-4-vl-9b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "llm-jp/llm-jp-4-vl-9b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "llm-jp/llm-jp-4-vl-9b",
		"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/llm-jp/llm-jp-4-vl-9b

SGLang

How to use llm-jp/llm-jp-4-vl-9b 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 "llm-jp/llm-jp-4-vl-9b" \
    --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": "llm-jp/llm-jp-4-vl-9b",
		"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 "llm-jp/llm-jp-4-vl-9b" \
        --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": "llm-jp/llm-jp-4-vl-9b",
		"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 llm-jp/llm-jp-4-vl-9b with Docker Model Runner:
```
docker model run hf.co/llm-jp/llm-jp-4-vl-9b
```

llm-jp-4-vl-9b / nemotron_subsets.md

speed

Update model card: eval tables + curve figure + Nemotron subset list

bf0c4e1 verified 23 days ago

preview code

Raw

History Blame Contribute Delete

937 Bytes

	# Nemotron-Image-Training-v3 subsets used

	For reproducibility, the following lists the exact subsets of
	[nvidia/Nemotron-Image-Training-v3](https://huggingface.co/datasets/nvidia/Nemotron-Image-Training-v3)
	used to train LLM-jp-4-VL 9B, after removing subsets with license / terms-of-use issues.

	\| subset \| samples \|
	\| :---- \| ----: \|
	\| openimages_1 \| 1,236,601 \|
	\| openimages_3 \| 1,236,572 \|
	\| cc3m \| 867,065 \|
	\| openimages_4 \| 504,275 \|
	\| openimages_2 \| 504,262 \|
	\| mulberry_2 \| 191,329 \|
	\| openimages_5 \| 92,905 \|
	\| ecd \| 55,341 \|
	\| long_document_arxiv_1 \| 53,990 \|
	\| long_document_arxiv_2 \| 46,247 \|
	\| docvqa \| 35,759 \|
	\| textvqa_commercial \| 34,602 \|
	\| long_document_arxiv_3 \| 31,089 \|
	\| textcaps \| 27,109 \|
	\| aokvqa_1 \| 19,200 \|
	\| zhwiki_2 \| 11,341 \|
	\| long_document_sec_3 \| 8,306 \|
	\| pubtables_1m \| 5,013 \|
	\| long_document_sec_4 \| 3,579 \|
	\| geometry3k \| 2,078 \|
	\| hiertext \| 514 \|
	\| turing \| 193 \|
	\| total \| 4,967,370 \|

	# Nemotron-Image-Training-v3 subsets used

	For reproducibility, the following lists the exact subsets of
	[nvidia/Nemotron-Image-Training-v3](https://huggingface.co/datasets/nvidia/Nemotron-Image-Training-v3)
	used to train LLM-jp-4-VL 9B, after removing subsets with license / terms-of-use issues.

	\| subset \| samples \|
	\| :---- \| ----: \|
	\| openimages_1 \| 1,236,601 \|
	\| openimages_3 \| 1,236,572 \|
	\| cc3m \| 867,065 \|
	\| openimages_4 \| 504,275 \|
	\| openimages_2 \| 504,262 \|
	\| mulberry_2 \| 191,329 \|
	\| openimages_5 \| 92,905 \|
	\| ecd \| 55,341 \|
	\| long_document_arxiv_1 \| 53,990 \|
	\| long_document_arxiv_2 \| 46,247 \|
	\| docvqa \| 35,759 \|
	\| textvqa_commercial \| 34,602 \|
	\| long_document_arxiv_3 \| 31,089 \|
	\| textcaps \| 27,109 \|
	\| aokvqa_1 \| 19,200 \|
	\| zhwiki_2 \| 11,341 \|
	\| long_document_sec_3 \| 8,306 \|
	\| pubtables_1m \| 5,013 \|
	\| long_document_sec_4 \| 3,579 \|
	\| geometry3k \| 2,078 \|
	\| hiertext \| 514 \|
	\| turing \| 193 \|
	\| total \| 4,967,370 \|