Instructions to use ByteDance/Sa2VA-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ByteDance/Sa2VA-4B with Transformers:

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

pipe = pipeline("image-text-to-text", model="ByteDance/Sa2VA-4B", trust_remote_code=True)
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 AutoModel
model = AutoModel.from_pretrained("ByteDance/Sa2VA-4B", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use ByteDance/Sa2VA-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ByteDance/Sa2VA-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ByteDance/Sa2VA-4B",
		"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/ByteDance/Sa2VA-4B

SGLang

How to use ByteDance/Sa2VA-4B 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 "ByteDance/Sa2VA-4B" \
    --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": "ByteDance/Sa2VA-4B",
		"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 "ByteDance/Sa2VA-4B" \
        --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": "ByteDance/Sa2VA-4B",
		"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 ByteDance/Sa2VA-4B with Docker Model Runner:
```
docker model run hf.co/ByteDance/Sa2VA-4B
```

Xiangtai commited on Jan 8, 2025

Commit

30715cb

verified ·

1 Parent(s): 68b1a30

Upload folder using huggingface_hub

Browse files

Files changed (1) hide show

README.md +16 -12

README.md CHANGED Viewed

@@ -3,9 +3,13 @@ license: mit
 pipeline_tag: image-text-to-text
 library_name: transformers
 base_model:
   - OpenGVLab/InternVL2_5-8B
   - OpenGVLab/InternViT-300M-448px-V2_5
   - internlm/internlm2_5-7b-chat
 base_model_relation: merge
 language:
   - multilingual
@@ -16,8 +20,8 @@ tags:
 # Sa2VA: Marrying SAM2 with LLaVA for Dense Grounded Understanding of Images and Videos
-[\[📂 GitHub\]](https://github.com/lxtGH/Sa2VA_opensource)
-[\[📜 Sa2VA paper\]]()
 [\[🚀 Quick Start\]](#quick-start)
@@ -30,11 +34,11 @@ Sa2VA is an MLLM capable of question answering, visual prompt understanding, and
 We built the Sa2VA series based on Qwen2-VL and InternVL2/2.5. In the following table, we provide some Sa2VA models built on InternVL2.5. Other Sa2VA models will be open-sourced soon.
-| Model Name |                             Base MLLM                             |                                 Language Part                                 |                       HF Link                        |
-|:----------:|:-----------------------------------------------------------------:|:-----------------------------------------------------------------------------:|:----------------------------------------------------:|
-|  Sa2VA-1B  | [InternVL2.0-1B](https://huggingface.co/OpenGVLab/InternVL2-1B) |   [Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct)    | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-1B) |
-|  Sa2VA-4B  | [InternVL2.5-4B](https://huggingface.co/OpenGVLab/InternVL2_5-4B) |    [Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)     | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-4B) |
-|  Sa2VA-8B  | [InternVL2.5-8B](https://huggingface.co/OpenGVLab/InternVL2_5-8B) |  [internlm2_5-7b-chat](https://huggingface.co/internlm/internlm2_5-7b-chat)   | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-8B) |
 ## Sa2VA Performance
 | Model Name |                             MMBench                             |                                    MME                                     |                       RefCOCO                        | RefCOCO+ | RefCOCOg | MeVIS | DAVIS | ReVOS |
@@ -67,7 +71,7 @@ tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast
 # for image chat
 image_path = "/PATH/TO/IMAGE"
-text_prompts = "<image>\nPlease describe the image."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
@@ -81,7 +85,7 @@ answer = return_dict["prediction"] # the text format answer
 # for image chat with segmentation output
 image_path = "/PATH/TO/IMAGE"
-text_prompts = "<image>\nCould you please give me a brief description of the image? Please respond with interleaved segmentation masks for the corresponding parts of the answer."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
@@ -97,7 +101,7 @@ masks = return_dict['prediction_masks']  # segmentation masks, list(np.array(1,
 # for chat with visual prompt (mask format) input
 mask_prompts = np.load('/PATH/TO/pred_masks.npy') # np.array(n_prompts, h, w)
 image_path = "/PATH/TO/IMAGE"
-text_prompts = "<image>\nCan you provide me with a detailed description of the region in the picture marked by region1."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
@@ -116,7 +120,7 @@ images_paths = [os.path.join(video_folder, image_path) for image_name in images_
 if len(images_paths) > 5:  # uniformly sample 5 frames
     step = (len(images_paths) - 1) // (5 - 1)
     images_paths = [images_paths[0]] + images_paths[1:-1][::step][1:] + [images_paths[-1]]
-text_prompts = "<image>\nPlease describe the video."
 input_dict = {
     'video': images_paths,
     'text': text_prompts,
@@ -132,7 +136,7 @@ answer = return_dict["prediction"] # the text format answer
 video_folder = "/PATH/TO/VIDEO_FOLDER"
 images_paths = os.listdir(video_folder)
 images_paths = [os.path.join(video_folder, image_path) for image_name in images_paths]
-text_prompts = "<image>\nPlease segment the person."
 input_dict = {
     'video': images_paths,
     'text': text_prompts,

 pipeline_tag: image-text-to-text
 library_name: transformers
 base_model:
+  - OpenGVLab/InternVL2-1B
   - OpenGVLab/InternVL2_5-8B
+  - OpenGVLab/InternVL2_5-4B
   - OpenGVLab/InternViT-300M-448px-V2_5
   - internlm/internlm2_5-7b-chat
+  - Qwen/Qwen2-0.5B-Instruct
+  - Qwen/Qwen2.5-3B-Instruct
 base_model_relation: merge
 language:
   - multilingual
 # Sa2VA: Marrying SAM2 with LLaVA for Dense Grounded Understanding of Images and Videos
+[\[📂 GitHub\]](https://github.com/magic-research/Sa2VA)
+[\[📜 Sa2VA paper\]](https://arxiv.org/abs/2501.04001)
 [\[🚀 Quick Start\]](#quick-start)
 We built the Sa2VA series based on Qwen2-VL and InternVL2/2.5. In the following table, we provide some Sa2VA models built on InternVL2.5. Other Sa2VA models will be open-sourced soon.
+| Model Name |                             Base MLLM                             |                                Language Part                                |                       HF Link                        |
+|:----------:|:-----------------------------------------------------------------:|:---------------------------------------------------------------------------:|:----------------------------------------------------:|
+|  Sa2VA-1B  | [InternVL2.0-1B](https://huggingface.co/OpenGVLab/InternVL2-1B) |   [Qwen2-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2-0.5B-Instruct)    | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-1B) |
+|  Sa2VA-4B  | [InternVL2.5-4B](https://huggingface.co/OpenGVLab/InternVL2_5-4B) |   [Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)    | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-4B) |
+|  Sa2VA-8B  | [InternVL2.5-8B](https://huggingface.co/OpenGVLab/InternVL2_5-8B) | [internlm2_5-7b-chat](https://huggingface.co/internlm/internlm2_5-7b-chat)  | [🤗 link](https://huggingface.co/ByteDance/Sa2VA-8B) |
 ## Sa2VA Performance
 | Model Name |                             MMBench                             |                                    MME                                     |                       RefCOCO                        | RefCOCO+ | RefCOCOg | MeVIS | DAVIS | ReVOS |
 # for image chat
 image_path = "/PATH/TO/IMAGE"
+text_prompts = "<image>Please describe the image."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
 # for image chat with segmentation output
 image_path = "/PATH/TO/IMAGE"
+text_prompts = "<image>Could you please give me a brief description of the image? Please respond with interleaved segmentation masks for the corresponding parts of the answer."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
 # for chat with visual prompt (mask format) input
 mask_prompts = np.load('/PATH/TO/pred_masks.npy') # np.array(n_prompts, h, w)
 image_path = "/PATH/TO/IMAGE"
+text_prompts = "<image>Can you provide me with a detailed description of the region in the picture marked by region1."
 image = Image.open(image_path).convert('RGB')
 input_dict = {
     'image': image,
 if len(images_paths) > 5:  # uniformly sample 5 frames
     step = (len(images_paths) - 1) // (5 - 1)
     images_paths = [images_paths[0]] + images_paths[1:-1][::step][1:] + [images_paths[-1]]
+text_prompts = "<image>Please describe the video."
 input_dict = {
     'video': images_paths,
     'text': text_prompts,
 video_folder = "/PATH/TO/VIDEO_FOLDER"
 images_paths = os.listdir(video_folder)
 images_paths = [os.path.join(video_folder, image_path) for image_name in images_paths]
+text_prompts = "<image>Please segment the person."
 input_dict = {
     'video': images_paths,
     'text': text_prompts,