Instructions to use ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only with Transformers:

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

pipe = pipeline("image-text-to-text", model="ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only")
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("ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only")
model = AutoModelForMultimodalLM.from_pretrained("ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only")
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 ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only",
		"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/ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only

SGLang

How to use ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only 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 "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only" \
    --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": "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only",
		"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 "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only" \
        --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": "ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only",
		"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 ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only with Docker Model Runner:
```
docker model run hf.co/ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only
```

Cosmos3 Nano Reasoner BNB8 VLM

This is an unofficial, locally repacked derivative of nvidia/Cosmos3-Nano for the VLM/reasoner path only.

It keeps the understanding tower used by the local vLLM loader:

model.language_model.*
lm_head.*
model.visual.*

It removes the Cosmos3 generator-side tensors, including diffusion *_moe_gen, added cross-attention projections, video/audio/action projection modules, VAE/sound-tokenizer assets, and Diffusers pipeline metadata.

The tensors are stored with stock Qwen3-VL key names and the config advertises Qwen3VLForConditionalGeneration, so consumers should not need the local vllm_cosmos3 plugin just to load the VLM path.

Original model materials are governed by the NVIDIA Open Model Dataset and Weights License 1.1. Retain upstream notices and review the original model card before redistribution or deployment.

Downloads last month: -

Safetensors

Model size

9B params

Tensor type

F32

BF16

Model tree for ThePyProgrammer/Cosmos3-Nano-reasoner-bnb8-vllm-und-only

Base model

nvidia/Cosmos3-Nano

Quantized

(9)

this model