Instructions to use Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact with Transformers:

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

pipe = pipeline("text-generation", model="Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact")
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("Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact")
model = AutoModelForMultimodalLM.from_pretrained("Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact", 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=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact

SGLang

How to use Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact 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 "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact" \
    --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": "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact",
		"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 "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact" \
        --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": "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact with Docker Model Runner:
```
docker model run hf.co/Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact
```

Gemma-4-31B-IT-NVFP4-24GB-compact

Compact NVFP4 quantized version of google/gemma-4-31B-it for vLLM.

Quantization Profile

Text MLP: NVFP4
Text self-attention: NVFP4
Token embeddings: NVFP4
lm_head: higher precision
Vision tower and vision embeddings: higher precision
KV cache: FP8

Usage

vllm serve Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact \
  --quantization modelopt \
  --gpu-memory-utilization 0.90

Official Gemma 4 vLLM recipe:

https://docs.vllm.ai/projects/recipes/en/latest/Google/Gemma4.html

Text Generation

curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact",
    "messages": [
      {"role": "user", "content": "Explain quantum entanglement in simple terms."}
    ],
    "max_tokens": 512
  }'

Downloads last month: 135

Safetensors

Model size

17B params

Tensor type

BF16

F8_E4M3

Model tree for Neural-ICE/Gemma-4-31B-IT-NVFP4-24GB-compact

Base model

google/gemma-4-31B

Finetuned

google/gemma-4-31B-it

Quantized

(310)

this model