Update README.md

README.md

@@ -31,29 +31,68 @@ The vision of OpenCSG is to empower every industry, every company, and every ind
 ## Model Description
 
 
+[CSG-VL](https://github.com/OpenCSGs/csg-vl) is a family of small but strong multimodal models. It offers multiple plug-and-play vision encoders, like EVA-CLIP, SigLIP and language backbones, including Wukong-1B, Llama-3-8B, Phi-1.5, StableLM-2, Qwen1.5 and Phi-2.
 
 
-**csg-wukong-1B-VL-v0.1** was finetuned on [csg-wukong-1B](https://huggingface.co/opencsg/csg-wukong-1B). 
-<br>
-we will introduce more information about this model.
+## Quickstart
 
-## Model Evaluation results
+Here we show a code snippet to show you how to use the model with transformers.
 
-We submitted csg-wukong-1B on the [open_llm_leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard), and
-the results show our model ranked the 8th among the ~1.5B pretrained small language models.
+Before running the snippet, you need to install the following dependencies:
 
+```shell
+pip install torch transformers accelerate pillow
+```
 
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/661790397437201d78141856/_HRTxL6N0qnNPNt-P8k9k.png)
+If the CUDA memory is enough, it would be faster to execute this snippet by setting `CUDA_VISIBLE_DEVICES=0`.
 
+Users especially those in Chinese mainland may want to refer to a HuggingFace [mirror site](https://hf-mirror.com). 
 
 
-# Training
+```python
+import torch
+import transformers
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from PIL import Image
+import warnings
 
-## Hardware
+# disable some warnings
+transformers.logging.set_verbosity_error()
+transformers.logging.disable_progress_bar()
+warnings.filterwarnings('ignore')
 
-- **GPUs:** 16 H800 
-- **Training time:** 43days 
+# set device
+torch.set_default_device('cpu')  # or 'cuda'
 
+model_name = 'opencsg/csg-wukong-1B-VL-v0.1'
+# create model
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16,
+    device_map='auto',
+    trust_remote_code=True)
+tokenizer = AutoTokenizer.from_pretrained(
+    model_name,
+    trust_remote_code=True)
+
+# text prompt
+prompt = 'What is the astronaut holding in his hand?'
+text = f"A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: <image>\n{prompt} ASSISTANT:"
+text_chunks = [tokenizer(chunk).input_ids for chunk in text.split('<image>')]
+input_ids = torch.tensor(text_chunks[0] + [-200] + text_chunks[1], dtype=torch.long).unsqueeze(0)
+image = Image.open('example_1.png')
+image_tensor = model.process_images([image], model.config).to(dtype=model.dtype)
+
+# generate
+output_ids = model.generate(
+    input_ids,
+    images=image_tensor,
+    max_new_tokens=100,
+    use_cache=True)[0]
+
+print(tokenizer.decode(output_ids[input_ids.shape[1]:], skip_special_tokens=True).strip())
+
+```
 ## Software
 
 - **Orchestration:** [Deepspeed](https://github.com/OpenCSGs)
@@ -88,22 +127,66 @@ OpenCSG的愿景是让每个行业、每个公司、每个人都拥有自己的
 ## 模型介绍
 
 
-**csg-wukong-1B-VL-v0.1** 在[csg-wukong-1B](https://huggingface.co/opencsg/csg-wukong-1B)预训练模型上微调而成.
-<br>
+[CSG-VL](https://github.com/OpenCSGs/csg-vl) 是一个小型但强大的多模式模型系列。它提供多种即插即用视觉编码器，如 EVA-CLIP、SigLIP 和语言主干，包括 Wukong-1B、Llama-3-8B、Phi-1.5、StableLM-2、Qwen1.5 和 Phi-2。
+
+
+## 快速开始
+下面我们将展示一个代码片段，告诉您如何使用带有转换器的模型。
+
+在运行该代码段之前，您需要安装以下依赖项：
+
+```shell
+pip install torch transformers accelerate pillow
+```
+
+如果 CUDA 内存足够，通过设置 CUDA_VISIBLE_DEVICES=0 来执行此代码片段会更快。
 
-我们将在后面介绍更多关于这个模型的信息。
+用户尤其是中国大陆的用户可能需要参考 HuggingFace [镜像站点](https://hf-mirror.com)。
 
 
-## 模型评测结果
+```python
+import torch
+import transformers
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from PIL import Image
+import warnings
 
-我们把csg-wukong-1B模型提交到[open_llm_leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)榜单上，结果显示我们的模型目前在~1.5B小语言模型中排名第8。
+# disable some warnings
+transformers.logging.set_verbosity_error()
+transformers.logging.disable_progress_bar()
+warnings.filterwarnings('ignore')
 
+# set device
+torch.set_default_device('cpu')  # or 'cuda'
 
-![image/png](https://cdn-uploads.huggingface.co/production/uploads/661790397437201d78141856/ZfWZ1Fd7ccKrJVx0okV9z.png)
+model_name = 'opencsg/csg-wukong-1B-VL-v0.1'
+# create model
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    torch_dtype=torch.float16,
+    device_map='auto',
+    trust_remote_code=True)
+tokenizer = AutoTokenizer.from_pretrained(
+    model_name,
+    trust_remote_code=True)
 
+# text prompt
+prompt = 'What is the astronaut holding in his hand?'
+text = f"A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: <image>\n{prompt} ASSISTANT:"
+text_chunks = [tokenizer(chunk).input_ids for chunk in text.split('<image>')]
+input_ids = torch.tensor(text_chunks[0] + [-200] + text_chunks[1], dtype=torch.long).unsqueeze(0)
+image = Image.open('example_1.png')
+image_tensor = model.process_images([image], model.config).to(dtype=model.dtype)
 
+# generate
+output_ids = model.generate(
+    input_ids,
+    images=image_tensor,
+    max_new_tokens=100,
+    use_cache=True)[0]
 
-# 训练
+print(tokenizer.decode(output_ids[input_ids.shape[1]:], skip_special_tokens=True).strip())
+```
 
 ## 硬件资源