Instructions to use webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4 with Transformers:

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

pipe = pipeline("image-text-to-text", model="webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4")
model = AutoModelForCausalLM.from_pretrained("webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4

SGLang

How to use webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4 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 "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4 with Docker Model Runner:
```
docker model run hf.co/webbrain-one/DeepSeek-V4-Flash-Vision-NVFP4
```

DeepSeek-V4-Flash-Vision-NVFP4 / hf_quant_config.json

webbrain-one

Copy hf_quant_config.json from pinned NVIDIA NVFP4 backbone (e3cd60e7)

bbbf7fe verified 6 days ago

Raw

History Blame Contribute Delete

4.55 kB

	{
	"producer": {
	"name": "modelopt",
	"version": "dsv4-nvfp4-experts"
	},
	"quantization": {
	"quant_algo": "MIXED_PRECISION",
	"kv_cache_quant_algo": null,
	"group_size": 16,
	"quantized_layers": {
	"layers.0.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.1.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.10.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.11.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.12.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.13.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.14.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.15.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.16.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.17.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.18.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.19.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.2.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.20.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.21.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.22.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.23.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.24.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.25.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.26.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.27.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.28.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.29.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.3.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.30.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.31.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.32.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.33.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.34.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.35.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.36.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.37.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.38.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.39.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.4.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.40.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.41.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.42.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.5.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.6.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.7.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.8.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	},
	"layers.9.ffn.experts": {
	"quant_algo": "NVFP4",
	"group_size": 16
	}
	},
	"exclude_modules": [
	".attn.",
	".ffn.shared_experts.",
	"head",
	"mtp.*"
	]
	}
	}