Instructions to use Qwen/Qwen2-VL-72B-Instruct-AWQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Local Apps Settings

How to use Qwen/Qwen2-VL-72B-Instruct-AWQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen2-VL-72B-Instruct-AWQ"
# 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/Qwen2-VL-72B-Instruct-AWQ",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen2-VL-72B-Instruct-AWQ

SGLang

How to use Qwen/Qwen2-VL-72B-Instruct-AWQ 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/Qwen2-VL-72B-Instruct-AWQ" \
    --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/Qwen2-VL-72B-Instruct-AWQ",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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/Qwen2-VL-72B-Instruct-AWQ" \
        --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/Qwen2-VL-72B-Instruct-AWQ",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen2-VL-72B-Instruct-AWQ with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen2-VL-72B-Instruct-AWQ
```

Error When trying to load the model with text-generation-inference

by mdpi-ai - opened Dec 2, 2024

Discussion

mdpi-ai

Dec 2, 2024

When trying to load this model using hugging face text-generation-inference docker image v1.4.2, I get the following error.

my parameters for TGI look like this
model_id: "Qwen/Qwen2-VL-72B-Instruct-AWQ"
num_shard: 1
cuda_memory_fraction: 1
max_top_n_tokens: 30
enable_cuda_graphs: true
cuda_visible_devices: 1
hf_token: ''
rope_scaling: 'dynamic'
rope_factor: 1
quantization: 'awq'

│ 370 │ │ """ │
│ ❱ 371 │ │ return self.weights_loader.get_weights_col_packed(self, prefix │
│ 372 │ │
│ 373 │ def get_weights_col(self, prefix: str): │
│ 374 │ │ return self.weights_loader.get_weights_col(self, prefix) │
│ │
│ ╭───────────────────────────────── locals ─────────────────────────────────╮ │
│ │ block_sizes = [16, 16, 16] │ │
│ │ prefix = 'visual.blocks.0.attn.qkv' │ │
│ │ self = <text_generation_server.utils.weights.Weights object at │ │
│ │ 0x7f8c7d7b4590> │ │
│ ╰──────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /opt/conda/lib/python3.11/site-packages/text_generation_server/layers/marlin │
│ /gptq.py:117 in get_weights_col_packed │
│ │
│ 114 │ │ │ │ f"{prefix}.qweight", dim=1, block_sizes=block_sizes │
│ 115 │ │ │ ) │
│ 116 │ │ except RuntimeError: │
│ ❱ 117 │ │ │ raise RuntimeError( │
│ 118 │ │ │ │ f"Cannot load {self.quantize} weight, make sure the │
│ 119 │ │ │ ) │
│ 120 │ │ scales = weights.get_packed_sharded( │
│ │
│ ╭───────────────────────────────── locals ─────────────────────────────────╮ │
│ │ block_sizes = [16, 16, 16] │ │
│ │ prefix = 'visual.blocks.0.attn.qkv' │ │
│ │ self = <text_generation_server.layers.marlin.gptq.GPTQMarlinWeig… │ │
│ │ object at 0x7f8c7d9238d0> │ │
│ │ weights = <text_generation_server.utils.weights.Weights object at │ │
│ │ 0x7f8c7d7b4590> │ │
│ ╰──────────────────────────────────────────────────────────────────────────╯ │
╰──────────────────────────────────────────────────────────────────────────────╯
RuntimeError: Cannot load awq weight, make sure the model is already
quantized. rank=0
2024-12-02T20:51:14.220341Z ERROR text_generation_launcher: Shard 0 failed to start

ArunSharma93

Mar 9, 2025

I get the same issue for the huggingface inference API for this model

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