Instructions to use Holy-fox/llm-jp-4-33b-thinking-nvfp4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Holy-fox/llm-jp-4-33b-thinking-nvfp4 with Transformers:

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

pipe = pipeline("text-generation", model="Holy-fox/llm-jp-4-33b-thinking-nvfp4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Holy-fox/llm-jp-4-33b-thinking-nvfp4")
model = AutoModelForCausalLM.from_pretrained("Holy-fox/llm-jp-4-33b-thinking-nvfp4", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.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(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Holy-fox/llm-jp-4-33b-thinking-nvfp4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Holy-fox/llm-jp-4-33b-thinking-nvfp4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Holy-fox/llm-jp-4-33b-thinking-nvfp4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Holy-fox/llm-jp-4-33b-thinking-nvfp4

SGLang

How to use Holy-fox/llm-jp-4-33b-thinking-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 "Holy-fox/llm-jp-4-33b-thinking-nvfp4" \
    --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": "Holy-fox/llm-jp-4-33b-thinking-nvfp4",
		"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 "Holy-fox/llm-jp-4-33b-thinking-nvfp4" \
        --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": "Holy-fox/llm-jp-4-33b-thinking-nvfp4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Holy-fox/llm-jp-4-33b-thinking-nvfp4 with Docker Model Runner:
```
docker model run hf.co/Holy-fox/llm-jp-4-33b-thinking-nvfp4
```

llm-jp-4-33b-thinking-nvfp4 / vllm_plugin /example_cli.py

Holy-fox

Add NVFP4 (W4A4) quantization of llm-jp-4-33b-thinking

76d60f4 verified 9 days ago

Raw

History Blame Contribute Delete

559 Bytes

	# This script works similarly to the `vllm` CLI command,
	# but registers additional components.
	#
	# Usage: python example_cli.py [...rest of vllm CLI arguments]
	#
	# Example:
	# The following command runs the llm-jp-4-8b-thinking model with the llmjp4 reasoning parser.
	# python example_cli.py serve llm-jp/llm-jp-4-8b-thinking --reasoning-parser llmjp4 --trust-remote-code

	from vllm.entrypoints.cli import main as cli_main

	# Load the custom reasoning before launching the CLI.
	import llmjp4_reasoning_parser


	if __name__ == "__main__":
	cli_main.main()