Instructions to use Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 with Transformers:

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

pipe = pipeline("text-generation", model="Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4")
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4")
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 Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4

SGLang

How to use Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 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 "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4" \
    --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": "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4",
		"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 "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4" \
        --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": "Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4
```

Int4为什么比没量化的float32和float16还慢

by hujianmin - opened Mar 8, 2025

Discussion

hujianmin

Mar 8, 2025

prompts =[
"A B C D E",
"one two three four five",
"When I was just a lad of ten, my father said to",
"Today I bought some",
"A majestic tiger walking through ",
"马龙是一名乒乓球",
"Imagine a breathtaking fantasy landscape during the golden hour, where the sun is setting behind a range of majestic snow-capped mountains. The sky is painted in vibrant hues of orange, pink, and purple, with scattered clouds reflecting the warm sunlight. In the foreground, a crystal-clear river winds through a lush valley, its surface shimmering with golden light. On the riverbank, a small medieval village with stone cottages and thatched roofs is nestled among blooming cherry blossom trees, their petals gently falling into the water. A cobblestone path leads from the village to a grand, ancient castle perched on a hill, surrounded by dense, enchanted forests with glowing mushrooms and ethereal blue fireflies.",
"中国的首都在",

]
对于这段prompts，new_token_length=32，int4耗时32s，float32和float16耗时为12s，设备时四块A100-40GB-PCIE

CHNtentes

Mar 24, 2025

用AutoGPTQ

wwf606

May 14, 2025

同样的问题，也是A100，int4的单专家forward时间大概是全精度的两倍

shing3232

May 15, 2025

同样的问题，也是A100，int4的单专家forward时间大概是全精度的两倍

这有什么奇怪的 gptq-int4是 W4A16的 int4需要dequant 成fp16才能decode

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment