Instructions to use Vegetabot/AVSQwen-Omni-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Vegetabot/AVSQwen-Omni-7B with Transformers:

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

pipe = pipeline("image-text-to-text", model="Vegetabot/AVSQwen-Omni-7B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForTextToWaveform

processor = AutoProcessor.from_pretrained("Vegetabot/AVSQwen-Omni-7B")
model = AutoModelForTextToWaveform.from_pretrained("Vegetabot/AVSQwen-Omni-7B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Vegetabot/AVSQwen-Omni-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Vegetabot/AVSQwen-Omni-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Vegetabot/AVSQwen-Omni-7B",
		"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/Vegetabot/AVSQwen-Omni-7B

SGLang

How to use Vegetabot/AVSQwen-Omni-7B 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 "Vegetabot/AVSQwen-Omni-7B" \
    --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": "Vegetabot/AVSQwen-Omni-7B",
		"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 "Vegetabot/AVSQwen-Omni-7B" \
        --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": "Vegetabot/AVSQwen-Omni-7B",
		"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 Vegetabot/AVSQwen-Omni-7B with Docker Model Runner:
```
docker model run hf.co/Vegetabot/AVSQwen-Omni-7B
```

AVSQwen-Omni-7B

GRPO post-trained Qwen2.5-Omni-7B for Referring Audio-Visual Segmentation (main model used for OmniAVS / RefAVS / MeViS / ReVOS / Ref-DAVIS / Ref-YouTube-VOS / ReasonSeg / RefCOCO). Trained on swift_train_bbox_grpo_balanced_full_v4_sam.jsonl with bbox_format + sam_keyframe + minimal_efficiency rewards.

Base model: Qwen/Qwen2.5-Omni-7B
Parameters: 7B
Training checkpoint: checkpoint-4000
Paper / code: https://github.com//AVSQwen

Usage

from transformers import Qwen2_5OmniForConditionalGeneration, Qwen2_5OmniProcessor

model = Qwen2_5OmniForConditionalGeneration.from_pretrained(
    "Vegetabot/AVSQwen-Omni-7B",
    torch_dtype="auto",
    device_map="auto",
)
processor = Qwen2_5OmniProcessor.from_pretrained("Vegetabot/AVSQwen-Omni-7B")

For the full inference pipeline (frame selector + grounding + SAM2 segmenter), please refer to inference/ and run/*.sh in the release repo.

Training

Framework: ms-swift with GRPO
Rewards: bbox_format_reward, sam_keyframe_reward, minimal_efficiency_reward
Data: swift_train_bbox_grpo_balanced_full_v4_sam.jsonl (OmniAVS + 7 referring datasets, balanced mix)
See training/train_qwen25omni_full_grpo_FINAL_V2.sh in the release repo.

Citation

@article{avsqwen2026,
  title   = {AVSQwen: ...},
  author  = {...},
  year    = {2026}
}

Downloads last month: 22

Safetensors

Model size

11B params

Tensor type

BF16

Model tree for Vegetabot/AVSQwen-Omni-7B

Base model

Qwen/Qwen2.5-Omni-7B

Finetuned

(53)

this model