Instructions to use tinnel123/OmniJev-0.8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tinnel123/OmniJev-0.8B with Transformers:

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

pipe = pipeline("image-text-to-text", model="tinnel123/OmniJev-0.8B")
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("tinnel123/OmniJev-0.8B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tinnel123/OmniJev-0.8B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tinnel123/OmniJev-0.8B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tinnel123/OmniJev-0.8B",
		"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/tinnel123/OmniJev-0.8B

SGLang

How to use tinnel123/OmniJev-0.8B 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 "tinnel123/OmniJev-0.8B" \
    --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": "tinnel123/OmniJev-0.8B",
		"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 "tinnel123/OmniJev-0.8B" \
        --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": "tinnel123/OmniJev-0.8B",
		"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 tinnel123/OmniJev-0.8B with Docker Model Runner:
```
docker model run hf.co/tinnel123/OmniJev-0.8B
```

OmniJev 0.8B · 全模态 Jev

Beijing Zhongguancun Academy · Institute of Automation, Chinese Academy of Sciences · Zevo

A decision layer on top of Qwen/Qwen3.5-0.8B that answers typed questions (choice / noul / score) about images, screens, game frames, video, sound (as spectrogram images) and robot views with calibrated probabilities in one forward pass, 0 generated tokens. Code and usage: https://github.com/tinnel123666888/OmniJev (from mso.infer import MSO1; MSO1("ckpt", "base") with this folder as ckpt and the backbone as base; the hybrid backbone is served through the branch path automatically).

Files

adapter_config.json / adapter_model.safetensors, head.pt (decision head), ord.pt (ordinal head), head_meta.json (head settings + per-type temperatures noul 0.87 / choice 0.94 / score 1.04), new_tok_emb.pt, processor and tokenizer files. ~123 MB.

Training data

Web operation (Mind2Web), grid pointing (web + phone), AndroidControl phone operation (10k steps, 96-cell tap grid), Simple-Wikipedia navigation, LIBERO-10 robot decisions, real-robot manipulation (UT-Austin MUTEX, 1,500 episodes), LIBERO-Spatial, Charades-STA video events, long video / planning / spatial families, a real-time catch game, Atari human play (11 games), Snake, Gomoku, chess from real games with local candidate moves, HaGRID gestures + fire/smoke/weapons, ESC-50 sounds rendered as spectrograms, a self-built mixed decision set (regions, existence, phone, chess, short video). About 270,000 decision records and 1.3 million typed questions, balanced across domains and identical for all three released sizes; packed multi-question training under proper scoring rules; temperatures fitted on held-out rows.

Held-out accuracy (serving path)

family	Qwen3.5-0.8B zero-shot	OmniJev 0.8B	OmniJev 4B (v0.5, Qwen3-VL)
LIBERO-10 robot decisions	0.551	0.771	0.821
Mind2Web test (task / website / domain)	0.401	0.596	0.770
Grid pointing, 96 cells (web)	0.384	0.474	0.729
Charades-STA video events	0.390	0.811	0.864
Catch game frames	0.409	0.661	0.885
HaGRID gestures + fire/smoke/weapons	0.529	0.969	0.974
OK-VQA answer pool	0.793	0.656	0.893
LongVideoBench val	0.410	0.486	0.588
Long video / planning / spatial	0.390	0.494	0.686
v0.1 image families	0.545	0.604	0.819
Wiki navigation	0.414	0.659	0.701
Real-robot MUTEX	0.442	0.681	–
Atari human play	–	0.707	–
Snake	–	0.727	–
Gomoku	–	0.694	–
Chess (local candidate moves)	–	0.603	–
AndroidControl phone operation	–	0.634	–
ESC-50 sounds (spectrogram)	–	0.473	–

Latency on one idle A800-40G (the serving path, a 768-token image budget, median of 12 runs with bench/speed_bench.py): one question in 216 ms, twelve questions about the same image in 236 ms, i.e. 19.6 ms per question. All three released sizes were measured back to back on that one card, so the numbers can be compared with each other.

License: Apache-2.0 for the weights and the code; the backbone keeps its own license.

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for tinnel123/OmniJev-0.8B

Base model

Qwen/Qwen3.5-0.8B-Base

Finetuned

Qwen/Qwen3.5-0.8B

Finetuned

(412)

this model