Instructions to use apus-ailab/APUS-OpenJev-v1-9B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use apus-ailab/APUS-OpenJev-v1-9B with Transformers:

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

pipe = pipeline("text-generation", model="apus-ailab/APUS-OpenJev-v1-9B")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("apus-ailab/APUS-OpenJev-v1-9B")
model = AutoModelForMultimodalLM.from_pretrained("apus-ailab/APUS-OpenJev-v1-9B", device_map="auto")
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 Settings

vLLM

How to use apus-ailab/APUS-OpenJev-v1-9B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "apus-ailab/APUS-OpenJev-v1-9B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "apus-ailab/APUS-OpenJev-v1-9B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/apus-ailab/APUS-OpenJev-v1-9B

SGLang

How to use apus-ailab/APUS-OpenJev-v1-9B 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 "apus-ailab/APUS-OpenJev-v1-9B" \
    --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": "apus-ailab/APUS-OpenJev-v1-9B",
		"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 "apus-ailab/APUS-OpenJev-v1-9B" \
        --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": "apus-ailab/APUS-OpenJev-v1-9B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use apus-ailab/APUS-OpenJev-v1-9B with Docker Model Runner:
```
docker model run hf.co/apus-ailab/APUS-OpenJev-v1-9B
```

APUS-OpenJev-v1-9B / README.zh-CN.md

gump2049

Add GGUF / MLX versions and Frozen80 results

82c9c56 verified 4 days ago

preview code

Raw

History Blame Contribute Delete

3.87 kB

metadata

library_name: transformers
license: apache-2.0
base_model: Qwen/Qwen3.5-9B
base_model_relation: finetune
pipeline_tag: text-generation
language:
  - en
  - zh
tags:
  - apus-openjev
  - decision-model
  - structured-output
  - bf16

APUS-OpenJev-v1-9B

English | 中文 · Collection · Model family · Technical Report · Runtime

基于 Qwen3.5 的通用决策模型，面向浏览器动作选择、业务流程路由与基于自然语言原则的判断。本仓库提供 9B checkpoint-3000 合并 BF16 权重，可独立下载，无需另外加载 LoRA。

技术特点

在当前请求提供的动态候选中打分与选择，返回完整候选分布。
配套原生运行时支持 effort=low（16层）与 effort=high（32层）；文本生成使用 high。
共享 Qwen 语言表示与词表投影；上层程序可将决策结果组装为结构化业务输出。

快速使用

python -m pip install huggingface_hub
hf download apus-ailab/APUS-OpenJev-v1-9B --local-dir ./APUS-OpenJev-v1-9B
cd APUS-OpenJev-v1-9B
python -m pip install -r requirements.txt
python examples.py . --device cuda:0 --effort high

评测与训练

本仓库的合并模型在 Frozen80 开发面板完整深度得分为 **68/80 (85.00%)**，覆盖 Browser、HelpSteer3、BoolQ、MNLI 和属性判断。详细结果见 merged-evaluation.json，训练与数据来源见 training.md。该面板用于工程回归，不能视为独立盲测或完整浏览器任务成功率。

候选分布是相对分数，未经校准不能当作正确率。合并为 BF16 后部分概率会变化，证据与限制见运行时指南。系列 9B 选择 checkpoint-3000，对应 85% 的开发面板成绩；checkpoint-5949 另保留在原系列库。

系列与下载

4B、9B、35B-A3B 按独立模型仓库发布，由 Collection 统一导航。每库公开下载数由 Hugging Face 统计；原系列仓库及旧路径保留。35B-A3B 的正式整模以独立验证和实际发布状态为准。

GGUF 与 MLX 版本

本模型（Frozen80 68/80）的量化版本，适用于 Ollama / llama.cpp / LM Studio 以及 Apple Silicon Mac：GGUF Collection · MLX Collection。

版本	Frozen80	与 BF16 决策一致
GGUF Q8_0	68/80	80/80
GGUF Q4_K_M	68/80	78/80
MLX 8bit	69/80	79/80
MLX 4bit	68/80	75/80

ollama run hf.co/apus-ailab/APUS-OpenJev-v1-9B-GGUF:Q8_0 --think=false

Ollama 必须关闭 thinking（--think=false 或 API 中 "think": false）。各版本的逐题评测和使用说明见对应仓库。

许可与致谢

感谢 Qwen/Qwen3.5-9B。权重及配套代码许可见 LICENSE 与来源说明。文件身份及上游固定版本见 release-manifest.json。

Authors: gumpcheng (xDAN2099), zhangxu, APUS AI-LAB.