Instructions to use BELLE-2/BELLE-VL with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use BELLE-2/BELLE-VL with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="BELLE-2/BELLE-VL", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("BELLE-2/BELLE-VL", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use BELLE-2/BELLE-VL with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "BELLE-2/BELLE-VL" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "BELLE-2/BELLE-VL", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/BELLE-2/BELLE-VL
- SGLang
How to use BELLE-2/BELLE-VL 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 "BELLE-2/BELLE-VL" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "BELLE-2/BELLE-VL", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "BELLE-2/BELLE-VL" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "BELLE-2/BELLE-VL", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use BELLE-2/BELLE-VL with Docker Model Runner:
docker model run hf.co/BELLE-2/BELLE-VL
Commit ·
106a9b1
1
Parent(s): 573ac83
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,22 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
+
|
| 5 |
+
## 📝Belle-VL
|
| 6 |
+
### 背景介绍
|
| 7 |
+
社区目前已经有很多多模态大语言模型相关开源工作,但大多以英文能力为主,比如[LLava](https://github.com/haotian-liu/LLaVA),[CogVLM](https://github.com/THUDM/CogVLM)等,而中文多模态大语言模型比如[VisualGLM-6B](https://github.com/THUDM/VisualGLM-6B)、[Qwen-VL](https://github.com/QwenLM/Qwen-VL)的语言模型基座均较小,实际应用中很难兼顾视觉和语言能力,因此Belle-VL选择基于更强的语言模型基座来扩展模型的视觉能力,为社区提供更加灵活的选择。
|
| 8 |
+
|
| 9 |
+
### 模型简介
|
| 10 |
+
在模型结构方面,我们主要参考的Qwen-VL 模型,原始Qwen-VL是基于Qwen7B模型训练而来,基座能力相对较弱,因此Belle-VL将语言模型扩展成了Qwen14B-chat,在中文语言能力和视觉能力方面可以兼顾,具备更好的扩展性。
|
| 11 |
+
|
| 12 |
+
### 训练策略
|
| 13 |
+
原始Qwen-vl采用了三阶段的训练方式,包括预训练、多任务训练和指令微调,依赖较大的数据和机器资源。受LLava1.5的启发,多模态指令微调比预训练更加重要,因此我们采用了两阶段的训练方式,如下图所示:
|
| 14 |
+

|
| 15 |
+
|
| 16 |
+
### 训练数据
|
| 17 |
+
* 预训练数据:预训练数据主要是基于LLava 的[558k](https://huggingface.co/datasets/liuhaotian/LLaVA-Pretrain)英文指令数据及其对应的中文翻译数据,此外我们还收集了[Flickr30k-CNA](https://zero.so.com/) 以及从[AI Challenger](https://tianchi.aliyun.com/dataset/145781?spm=a2c22.12282016.0.0.5c823721PG2nBW)随机选取的100k数据
|
| 18 |
+
|
| 19 |
+
* 多模态指令数据:指令微调阶段,数据主要来自[LLava](https://github.com/haotian-liu/LLaVA), [LRV-Instruction](https://github.com/FuxiaoLiu/LRV-Instruction), [LLaVAR](https://github.com/SALT-NLP/LLaVAR),[LVIS-INSTRUCT4V](https://github.com/X2FD/LVIS-INSTRUCT4V)等开源项目,我们也对其中部分数据进行了翻译,在此真诚的感谢他们为开源所做出的贡献!
|
| 20 |
+
|
| 21 |
+
### [MME Benchmark](https://github.com/BradyFU/Awesome-Multimodal-Large-Language-Models/tree/Evaluation)
|
| 22 |
+
待更新
|