Instructions to use tencent/Hunyuan-4B-Instruct-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tencent/Hunyuan-4B-Instruct-FP8 with Transformers:

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

pipe = pipeline("text-generation", model="tencent/Hunyuan-4B-Instruct-FP8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tencent/Hunyuan-4B-Instruct-FP8")
model = AutoModelForCausalLM.from_pretrained("tencent/Hunyuan-4B-Instruct-FP8")
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

vLLM

How to use tencent/Hunyuan-4B-Instruct-FP8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tencent/Hunyuan-4B-Instruct-FP8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tencent/Hunyuan-4B-Instruct-FP8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tencent/Hunyuan-4B-Instruct-FP8

SGLang

How to use tencent/Hunyuan-4B-Instruct-FP8 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 "tencent/Hunyuan-4B-Instruct-FP8" \
    --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": "tencent/Hunyuan-4B-Instruct-FP8",
		"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 "tencent/Hunyuan-4B-Instruct-FP8" \
        --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": "tencent/Hunyuan-4B-Instruct-FP8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tencent/Hunyuan-4B-Instruct-FP8 with Docker Model Runner:
```
docker model run hf.co/tencent/Hunyuan-4B-Instruct-FP8
```

manaestras commited on Jul 31, 2025

Commit

2d252b7

verified ·

1 Parent(s): d868535

Upload angelslim_config.json with huggingface_hub

Browse files

Files changed (1) hide show

angelslim_config.json +5 -5

angelslim_config.json CHANGED Viewed

@@ -11,7 +11,7 @@
     "compression_config": {
         "name": "PTQ",
         "quantization": {
-            "name": "fp8_static",
             "bits": 8,
             "quant_method": {
                 "weight": "per-tensor",
@@ -32,9 +32,9 @@
     },
     "dataset_config": {
         "name": "TextDataset",
-        "data_path": "./dataset/PTQ_data.jsonl",
         "max_seq_length": 4096,
-        "num_samples": 512,
         "batch_size": 1,
         "shuffle": false
     },
@@ -49,8 +49,8 @@
         "python": "3.10.14 (main, Mar  6 2025, 18:16:47) [GCC 11.4.0]",
         "angelslim": {
             "name": "angelslim",
-            "version": "be35620eca8ad4e6bf8263e84cee69d90c7045a4",
-            "source": "git"
         },
         "torch": {
             "name": "torch",

     "compression_config": {
         "name": "PTQ",
         "quantization": {
+            "name": "fp8_psad",
             "bits": 8,
             "quant_method": {
                 "weight": "per-tensor",
     },
     "dataset_config": {
         "name": "TextDataset",
+        "data_path": "/",
         "max_seq_length": 4096,
+        "num_samples": 32,
         "batch_size": 1,
         "shuffle": false
     },
         "python": "3.10.14 (main, Mar  6 2025, 18:16:47) [GCC 11.4.0]",
         "angelslim": {
             "name": "angelslim",
+            "version": "N/A",
+            "source": "Unknown"
         },
         "torch": {
             "name": "torch",