thanks to Qwen ❤

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+SimSun.ttf filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,666 @@
+---
+language:
+- zh
+- en
+tags:
+- qwen
+pipeline_tag: text-generation
+inference: false
+---
+
+# Qwen-VL-Chat-Int4
+
+<br>
+
+<p align="center">
+    <img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/logo_vl.jpg" width="400"/>
+<p>
+<br>
+
+<p align="center">
+        Qwen-VL <a href="https://modelscope.cn/models/qwen/Qwen-VL/summary">🤖 <a> | <a href="https://huggingface.co/Qwen/Qwen-VL">🤗</a>&nbsp ｜ Qwen-VL-Chat <a href="https://modelscope.cn/models/qwen/Qwen-VL-Chat/summary">🤖 <a>| <a href="https://huggingface.co/Qwen/Qwen-VL-Chat">🤗</a>&nbsp ｜ Qwen-VL-Chat-Int4 <a href="https://huggingface.co/Qwen/Qwen-VL-Chat-Int4">🤗</a>
+<br>
+<a href="assets/wechat.png">WeChat</a>&nbsp&nbsp | &nbsp&nbsp<a href="https://discord.gg/z3GAxXZ9Ce">Discord</a>&nbsp&nbsp | &nbsp&nbsp<a href="https://modelscope.cn/studios/qwen/Qwen-VL-Chat-Demo/summary">Demo</a>&nbsp ｜ &nbsp<a href="https://arxiv.org/abs/2308.12966">Report</a>
+</p>
+<br>
+
+**Qwen-VL** 是阿里云研发的大规模视觉语言模型（Large Vision Language Model, LVLM）。Qwen-VL 可以以图像、文本、检测框作为输入，并以文本和检测框作为输出。Qwen-VL 系列模型性能强大，具备多语言对话、多图交错对话等能力，并支持中文开放域定位和细粒度图像识别与理解。
+
+**Qwen-VL** (Qwen Large Vision Language Model) is the visual multimodal version of the large model series, Qwen (abbr. Tongyi Qianwen), proposed by Alibaba Cloud. Qwen-VL accepts image, text, and bounding box as inputs, outputs text and bounding box. The features of Qwen-VL include:
+
+目前，我们提供了Qwen-VL和Qwen-VL-Chat两个模型，分别为预训练模型和Chat模型。如果想了解更多关于模型的信息，请点击[链接](https://github.com/QwenLM/Qwen-VL/blob/master/visual_memo.md)查看我们的技术备忘录。本仓库为Qwen-VL-Chat的量化模型Qwen-VL-Chat-Int4仓库。
+
+We release Qwen-VL and Qwen-VL-Chat, which are pretrained model and Chat model respectively. For more details about Qwen-VL, please refer to our [technical memo](https://github.com/QwenLM/Qwen-VL/blob/master/visual_memo.md). This repo is the one for Qwen-VL-Chat-Int4.
+<br>
+
+## 安装要求 (Requirements)
+
+* python 3.8及以上版本
+* pytorch2.0及以上版本
+* 建议使用CUDA 11.4及以上
+* python 3.8 and above
+* pytorch 2.0 and above are recommended
+* CUDA 11.4 and above are recommended
+  <br>
+
+## 快速开始 (Quickstart)
+
+我们提供简单的示例来说明如何利用 🤗 Transformers 快速使用Qwen-VL-Chat-Int4。
+
+在开始前，请确保你已经配置好环境并安装好相关的代码包。最重要的是，确保你满足上述要求，然后安装相关的依赖库。
+
+Below, we provide simple examples to show how to use Qwen-VL-Chat-Int4 with 🤗 Transformers.
+
+Before running the code, make sure you have setup the environment and installed the required packages. Make sure you meet the above requirements, and then install the dependent libraries.
+
+```bash
+pip install -r requirements.txt
+pip install optimum
+
+git clone https://github.com/JustinLin610/AutoGPTQ.git & cd AutoGPTQ
+pip install -v .
+```
+
+接下来你可以开始使用Transformers来使用我们的模型。关于视觉模块的更多用法，请参考[教程](TUTORIAL.md)。
+
+Now you can start with Transformers. More usage aboue vision encoder, please refer to [tutorial](TUTORIAL_zh.md).
+
+#### 🤗 Transformers
+
+To use Qwen-VL-Chat-Int4 for the inference, all you need to do is to input a few lines of codes as demonstrated below. However, **please make sure that you are using the latest code.**
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+torch.manual_seed(1234)
+
+# Note: The default behavior now has injection attack prevention off.
+tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen-VL-Chat-Int4", trust_remote_code=True)
+
+# use cuda device
+model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen-VL-Chat-Int4", device_map="cuda", trust_remote_code=True).eval()
+
+# 1st dialogue turn
+query = tokenizer.from_list_format([
+    {'image': 'https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg'},
+    {'text': '这是什么'},
+])
+response, history = model.chat(tokenizer, query=query, history=None)
+print(response)
+# 图中是一名年轻女子在沙滩上和她的狗玩耍，狗的品种可能是拉布拉多。她们坐在沙滩上，狗的前腿抬起来，似乎在和人类击掌。两人之间充满了信任和爱。
+
+# 2nd dialogue turn
+response, history = model.chat(tokenizer, '输出"击掌"的检测框', history=history)
+print(response)
+# <ref>击掌</ref><box>(517,508),(589,611)</box>
+image = tokenizer.draw_bbox_on_latest_picture(response, history)
+if image:
+  image.save('1.jpg')
+else:
+  print("no box")
+```
+
+<p align="center">
+    <img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo_highfive.jpg" width="500"/>
+<p>
+<br>
+
+## 量化 (Quantization)
+
+### 效果评测 (Performance)
+
+我们列出不同精度下模型在评测基准 **[TouchStone](https://github.com/OFA-Sys/TouchStone)** 上的表现，并发现量化模型并没有显著性能损失。结果如下所示：
+
+We illustrate the model performance of both BF16 and Int4 models on the benchmark **[TouchStone](https://github.com/OFA-Sys/TouchStone)**, and we find that the quantized model does not suffer from significant performance degradation. Results are shown below:
+
+| Quantization | ZH.        | EN            |
+| ------------ | :--------: | :-----------: | 
+| BF16         | 401.2      |    645.2      |
+| Int4         | 386.6      |    651.4      |
+
+### 推理速度 (Inference Speed)
+
+我们测算了在输入一张图片（即258个token）的条件下BF16和Int4的模型生成1792 (2048-258) 和 7934 (8192-258) 个token的平均速度。
+
+We measured the average inference speed (tokens/s) of generating 1792 (2048-258) and 7934 (8192-258) tokens with the context of an image (which takes 258 tokens) under BF16 precision and Int4 quantization, respectively.
+
+| Quantization | Speed (2048 tokens) | Speed (8192 tokens) |
+| ------------ | :-----------------: | :-----------------: |
+| BF16         |        28.87        |        24.32        |
+| Int4         |        37.79        |        34.34        |
+
+推理速度测算是在单卡 A100-SXM4-80G GPU上运行，使用PyTorch 2.0.1及CUDA 11.4。
+
+The profiling runs on a single A100-SXM4-80G GPU with PyTorch 2.0.1 and CUDA 11.4.
+
+### GPU显存占用 (GPU Memory Usage)
+
+我们还测算了在一张图片输入的条件下BF16和Int4模型生成1792 (2048-258) 和 7934 (8192-258) 个token所需显存。结果如下所示：
+
+We also profile the peak GPU memory usage for encoding 1792 (2048-258) tokens (including an image) as context (and generating single token) and generating 7934 (8192-258) tokens (with an image as context) under BF16 or Int4 quantization level, respectively. The results are shown below.
+
+| Quantization | Peak Usage for Encoding 2048 Tokens | Peak Usage for Generating 8192 Tokens |
+| ------------ | :---------------------------------: | :-----------------------------------: |
+| BF16         |               22.60GB               |                28.01GB                |
+| Int4         |               11.82GB               |                17.23GB                |
+
+上述速度和显存测算使用[此脚本](https://qianwen-res.oss-cn-beijing.aliyuncs.com/profile_mm.py)完成。
+
+The above speed and memory profiling are conducted using [this script](https://qianwen-res.oss-cn-beijing.aliyuncs.com/profile_mm.py).
+<br>
+
+## 评测
+
+我们从两个角度评测了两个模型的能力：
+
+1. 在**英文标准 Benchmark** 上评测模型的基础任务能力。目前评测了四大类多模态任务：
+   
+   - Zero-shot Caption: 评测模型在未见过数据集上的零样本图片描述能力；
+   - General VQA: 评测模型的通用问答能力，例如判断题、颜色、个数、类目等问答能力；
+   - Text-based VQA：评测模型对于图片中文字相关的识别/问答能力，例如文档问答、图表问答、文字问答等；
+   - Referring Expression Compression：评测模型给定物体描述画检测框的能力；
+2. **试金石 (TouchStone)**：为了评测模型整体的图文对话能力和人类对齐水平。我们为此构建了一个基于 GPT4 打分来评测 LVLM 模型的 Benchmark：TouchStone。在 TouchStone-v0.1 中：
+   
+   - 评测基准总计涵盖 300+张图片、800+道题目、27个类别。包括基础属性问答、人物地标问答、影视作品问答、视觉推理、反事实推理、诗歌创作、故事写作，商品比较、图片解题等**尽可能广泛的类别**。
+   - 为了弥补目前 GPT4 无法直接读取图片的缺陷，我们给所有的带评测图片提供了**人工标注的充分详细描述**，并且将图片的详细描述、问题和模型的输出结果一起交给 GPT4 打分。
+   - 评测同时包含英文版本和中文版本。
+
+评测结果如下：
+
+We evaluated the model's ability from two perspectives:
+
+1. **Standard Benchmarks**: We evaluate the model's basic task capabilities on four major categories of multimodal tasks:
+   
+   - Zero-shot Caption: Evaluate model's zero-shot image captioning ability on unseen datasets;
+   - General VQA: Evaluate the general question-answering ability of pictures, such as the judgment, color, number, category, etc;
+   - Text-based VQA: Evaluate the model's ability to recognize text in pictures, such as document QA, chart QA, etc;
+   - Referring Expression Comprehension: Evaluate the ability to localize a target object in an image described by a referring expression.
+2. **TouchStone**: To evaluate the overall text-image dialogue capability and alignment level with humans, we have constructed a benchmark called TouchStone, which is based on scoring with GPT4 to evaluate the LVLM model.
+   
+   - The TouchStone benchmark covers a total of 300+ images, 800+ questions, and 27 categories. Such as attribute-based Q&A, celebrity recognition, writing poetry, summarizing multiple images, product comparison, math problem solving, etc;
+   - In order to break the current limitation of GPT4 in terms of direct image input, TouchStone provides fine-grained image annotations by human labeling. These detailed annotations, along with the questions and the model's output, are then presented to GPT4 for scoring.
+   - The benchmark includes both English and Chinese versions.
+
+The results of the evaluation are as follows:
+
+Qwen-VL outperforms current SOTA generalist models on multiple VL tasks and has a more comprehensive coverage in terms of capability range.
+
+<p align="center">
+    <img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/radar.png" width="600"/>
+<p>
+
+### 零样本图像描述 & 通用视觉问答 (Zero-shot Captioning & General VQA)
+
+<table>
+<thead>
+  <tr>
+    <th rowspan="2">Model type</th>
+    <th rowspan="2">Model</th>
+    <th colspan="2">Zero-shot Captioning</th>
+    <th colspan="5">General VQA</th>
+  </tr>
+  <tr>
+    <th>NoCaps</th>
+    <th>Flickr30K</th>
+    <th>VQAv2<sup>dev</sup></th>
+    <th>OK-VQA</th>
+    <th>GQA</th>
+    <th>SciQA-Img<br>(0-shot)</th>
+    <th>VizWiz<br>(0-shot)</th>
+  </tr>
+</thead>
+<tbody align="center">
+  <tr>
+    <td rowspan="10">Generalist<br>Models</td>
+    <td>Flamingo-9B</td>
+    <td>-</td>
+    <td>61.5</td>
+    <td>51.8</td>
+    <td>44.7</td>
+    <td>-</td>
+    <td>-</td>
+    <td>28.8</td>
+  </tr>
+  <tr>
+    <td>Flamingo-80B</td>
+    <td>-</td>
+    <td>67.2</td>
+    <td>56.3</td>
+    <td>50.6</td>
+    <td>-</td>
+    <td>-</td>
+    <td>31.6</td>
+  </tr>
+  <tr>
+    <td>Unified-IO-XL</td>
+    <td>100.0</td>
+    <td>-</td>
+    <td>77.9</td>
+    <td>54.0</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>Kosmos-1</td>
+    <td>-</td>
+    <td>67.1</td>
+    <td>51.0</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>29.2</td>
+  </tr>
+  <tr>
+    <td>Kosmos-2</td>
+    <td>-</td>
+    <td>66.7</td>
+    <td>45.6</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>BLIP-2 (Vicuna-13B)</td>
+    <td>103.9</td>
+    <td>71.6</td>
+    <td>65.0</td>
+    <td>45.9</td>
+    <td>32.3</td>
+    <td>61.0</td>
+    <td>19.6</td>
+  </tr>
+  <tr>
+    <td>InstructBLIP (Vicuna-13B)</td>
+    <td><strong>121.9</strong></td>
+    <td>82.8</td>
+    <td>-</td>
+    <td>-</td>
+    <td>49.5</td>
+    <td>63.1</td>
+    <td>33.4</td>
+  </tr>
+  <tr>
+    <td>Shikra (Vicuna-13B)</td>
+    <td>-</td>
+    <td>73.9</td>
+    <td>77.36</td>
+    <td>47.16</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td><strong>Qwen-VL (Qwen-7B)</strong></td>
+    <td>121.4</td>
+    <td><b>85.8</b></td>
+    <td><b>78.8</b></td>
+    <td><b>58.6</b></td>
+    <td><b>59.3</b></td>
+    <td>67.1</td>
+    <td>35.2</td>
+  </tr>
+  <!-- <tr>
+    <td>Qwen-VL (4-shot)</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>63.6</td>
+    <td>-</td>
+    <td>-</td>
+    <td>39.1</td>
+  </tr> -->
+  <tr>
+    <td>Qwen-VL-Chat</td>
+    <td>120.2</td>
+    <td>81.0</td>
+    <td>78.2</td>
+    <td>56.6</td>
+    <td>57.5</td>
+    <td><b>68.2</b></td>
+    <td><b>38.9</b></td>
+  </tr>
+  <!-- <tr>
+    <td>Qwen-VL-Chat (4-shot)</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>60.6</td>
+    <td>-</td>
+    <td>-</td>
+    <td>44.45</td>
+  </tr> -->
+  <tr>
+    <td>Previous SOTA<br>(Per Task Fine-tuning)</td>
+    <td>-</td>
+    <td>127.0<br>(PALI-17B)</td>
+    <td>84.5<br>(InstructBLIP<br>-FlanT5-XL)</td>
+    <td>86.1<br>(PALI-X<br>-55B)</td>
+    <td>66.1<br>(PALI-X<br>-55B)</td>
+    <td>72.1<br>(CFR)</td>
+    <td>92.53<br>(LLaVa+<br>GPT-4)</td>
+    <td>70.9<br>(PALI-X<br>-55B)</td>
+  </tr>
+</tbody>
+</table>
+
+- 在 Zero-shot Caption 中，Qwen-VL 在 Flickr30K 数据集上取得了 **SOTA** 的结果，并在 Nocaps 数据集上取得了和 InstructBlip 可竞争的结果。
+- 在 General VQA 中，Qwen-VL 取得了 LVLM 模型同等量级和设定下 **SOTA** 的结果。
+- For zero-shot image captioning, Qwen-VL achieves the **SOTA** on Flickr30K and competitive results on Nocaps with InstructBlip.
+- For general VQA, Qwen-VL achieves the **SOTA** under the same generalist LVLM scale settings.
+
+### 文本导向的视觉问答 (Text-oriented VQA)
+
+<table>
+<thead>
+  <tr>
+    <th>Model type</th>
+    <th>Model</th>
+    <th>TextVQA</th>
+    <th>DocVQA</th>
+    <th>ChartQA</th>
+    <th>AI2D</th>
+    <th>OCR-VQA</th>
+  </tr>
+</thead>
+<tbody align="center">
+  <tr>
+    <td rowspan="5">Generalist Models</td>
+    <td>BLIP-2 (Vicuna-13B)</td>
+    <td>42.4</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>InstructBLIP (Vicuna-13B)</td>
+    <td>50.7</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>mPLUG-DocOwl (LLaMA-7B)</td>
+    <td>52.6</td>
+    <td>62.2</td>
+    <td>57.4</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>Pic2Struct-Large (1.3B)</td>
+    <td>-</td>
+    <td><b>76.6</b></td>
+    <td>58.6</td>
+    <td>42.1</td>
+    <td>71.3</td>
+  </tr>
+  <tr>
+    <td>Qwen-VL (Qwen-7B)</td>
+    <td><b>63.8</b></td>
+    <td>65.1</td>
+    <td><b>65.7</b></td>
+    <td><b>62.3</b></td>
+    <td><b>75.7</b></td>
+  </tr>
+  <tr>
+    <td>Specialist SOTAs<br>(Specialist/Finetuned)</td>
+    <td>PALI-X-55B (Single-task FT)<br>(Without OCR Pipeline)</td>
+    <td>71.44</td>
+    <td>80.0</td>
+    <td>70.0</td>
+    <td>81.2</td>
+    <td>75.0</td>
+  </tr>
+</tbody>
+</table>
+
+- 在文字相关的识别/问答评测上，取得了当前规模下通用 LVLM 达到的最好结果。
+- 分辨率对上述某几个评测非常重要，大部分 224 分辨率的开源 LVLM 模型无法完成以上评测，或只能通过切图的方式解决。Qwen-VL 将分辨率提升到 448，可以直接以端到端的方式进行以上评测。Qwen-VL 在很多任务上甚至超过了 1024 分辨率的 Pic2Struct-Large 模型。
+- In text-related recognition/QA evaluation, Qwen-VL achieves the SOTA under the generalist LVLM scale settings.
+- Resolution is important for several above evaluations. While most open-source LVLM models with 224 resolution are incapable of these evaluations or can only solve these by cutting images, Qwen-VL scales the resolution to 448 so that it can be evaluated end-to-end. Qwen-VL even outperforms Pic2Struct-Large models of 1024 resolution on some tasks.
+
+### 细粒度视觉定位 (Referring Expression Comprehension)
+
+<table>
+<thead>
+  <tr>
+    <th rowspan="2">Model type</th>
+    <th rowspan="2">Model</th>
+    <th colspan="3">RefCOCO</th>
+    <th colspan="3">RefCOCO+</th>
+    <th colspan="2">RefCOCOg</th>
+    <th>GRIT</th>
+  </tr>
+  <tr>
+    <th>val</th>
+    <th>test-A</th>
+    <th>test-B</th>
+    <th>val</th>
+    <th>test-A</th>
+    <th>test-B</th>
+    <th>val-u</th>
+    <th>test-u</th>
+    <th>refexp</th>
+  </tr>
+</thead>
+<tbody align="center">
+  <tr>
+    <td rowspan="8">Generalist Models</td>
+    <td>GPV-2</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>51.50</td>
+  </tr>
+  <tr>
+    <td>OFA-L*</td>
+    <td>79.96</td>
+    <td>83.67</td>
+    <td>76.39</td>
+    <td>68.29</td>
+    <td>76.00</td>
+    <td>61.75</td>
+    <td>67.57</td>
+    <td>67.58</td>
+    <td>61.70</td>
+  </tr>
+  <tr>
+    <td>Unified-IO</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td><b>78.61</b></td>
+  </tr>
+  <tr>
+    <td>VisionLLM-H</td>
+    <td></td>
+    <td>86.70</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>Shikra-7B</td>
+    <td>87.01</td>
+    <td>90.61</td>
+    <td>80.24 </td>
+    <td>81.60</td>
+    <td>87.36</td>
+    <td>72.12</td>
+    <td>82.27</td>
+    <td>82.19</td>
+    <td>69.34</td>
+  </tr>
+  <tr>
+    <td>Shikra-13B</td>
+    <td>87.83 </td>
+    <td>91.11</td>
+    <td>81.81</td>
+    <td>82.89</td>
+    <td>87.79</td>
+    <td>74.41</td>
+    <td>82.64</td>
+    <td>83.16</td>
+    <td>69.03</td>
+  </tr>
+  <tr>
+    <td>Qwen-VL-7B</td>
+    <td><b>89.36</b></td>
+    <td>92.26</td>
+    <td><b>85.34</b></td>
+    <td><b>83.12</b></td>
+    <td>88.25</td>
+    <td><b>77.21</b></td>
+    <td>85.58</td>
+    <td>85.48</td>
+    <td>78.22</td>
+  </tr>
+  <tr>
+    <td>Qwen-VL-7B-Chat</td>
+    <td>88.55</td>
+    <td><b>92.27</b></td>
+    <td>84.51</td>
+    <td>82.82</td>
+    <td><b>88.59</b></td>
+    <td>76.79</td>
+    <td><b>85.96</b></td>
+    <td><b>86.32</b></td>
+    <td>-</td>
+  <tr>
+    <td rowspan="3">Specialist SOTAs<br>(Specialist/Finetuned)</td>
+    <td>G-DINO-L</td>
+    <td>90.56&nbsp;&nbsp;</td>
+    <td>93.19</td>
+    <td>88.24</td>
+    <td>82.75</td>
+    <td>88.95</td>
+    <td>75.92</td>
+    <td>86.13</td>
+    <td>87.02</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>UNINEXT-H</td>
+    <td>92.64 </td>
+    <td>94.33</td>
+    <td>91.46</td>
+    <td>85.24</td>
+    <td>89.63</td>
+    <td>79.79</td>
+    <td>88.73</td>
+    <td>89.37</td>
+    <td>-</td>
+  </tr>
+  <tr>
+    <td>ONE-PEACE</td>
+    <td>92.58 </td>
+    <td>94.18</td>
+    <td>89.26</td>
+    <td>88.77</td>
+    <td>92.21</td>
+    <td>83.23</td>
+    <td>89.22</td>
+    <td>89.27</td>
+    <td>-</td>
+  </tr>
+</tbody>
+</table>
+
+- 在定位任务上，Qwen-VL 全面超过 Shikra-13B，取得了目前 Generalist LVLM 模型上在 Refcoco 上的 **SOTA**。
+- Qwen-VL 并没有在任何中文定位数据上训练过，但通过中文 Caption 数据和 英文 Grounding 数据的训练，可以 Zero-shot 泛化出中文 Grounding 能力。
+
+我们提供了以上**所有**评测脚本以供复现我们的实验结果。请阅读 [eval/EVALUATION.md](eval/EVALUATION.md) 了解更多信息。
+
+- Qwen-VL achieves the **SOTA** in all above referring expression comprehension benchmarks.
+- Qwen-VL has not been trained on any Chinese grounding data, but it can still generalize to the Chinese Grounding tasks in a zero-shot way by training Chinese Caption data and English Grounding data.
+
+We provide all of the above evaluation scripts for reproducing our experimental results. Please read [eval/EVALUATION.md](eval/EVALUATION.md) for more information.
+
+### 闲聊能力测评 (Chat Evaluation)
+
+TouchStone 是一个基于 GPT4 打分来评测 LVLM 模型的图文对话能力和人类对齐水平的基准。它涵盖了 300+张图片、800+道题目、27个类别，包括基础属性、人物地标、视觉推理、诗歌创作、故事写作、商品比较、图片解题等**尽可能广泛的类别**。关于 TouchStone 的详细介绍，请参考[touchstone/README_CN.md](touchstone/README_CN.md)了解更多信息。
+
+TouchStone is a benchmark based on scoring with GPT4 to evaluate the abilities of the LVLM model on text-image dialogue and alignment levels with humans. It covers a total of 300+ images, 800+ questions, and 27 categories, such as attribute-based Q&A, celebrity recognition, writing poetry, summarizing multiple images, product comparison, math problem solving, etc. Please read [touchstone/README_CN.md](touchstone/README.md) for more information.
+
+#### 英语 (English)
+
+| Model         | Score |
+|---------------|-------|
+| PandaGPT      | 488.5 |
+| MiniGPT4      | 531.7 |
+| InstructBLIP  | 552.4 |
+| LLaMA-AdapterV2 | 590.1 |
+| mPLUG-Owl     | 605.4 |
+| LLaVA         | 602.7 |
+| Qwen-VL-Chat   | 645.2 |
+
+#### 中文 (Chinese)
+
+| Model         | Score |
+|---------------|-------|
+| VisualGLM     | 247.1 |
+| Qwen-VL-Chat   | 401.2 |
+
+Qwen-VL-Chat 模型在中英文的对齐评测中均取得当前 LVLM 模型下的最好结果。
+
+Qwen-VL-Chat has achieved the best results in both Chinese and English alignment evaluation.
+<br>
+
+## 常见问题 (FAQ)
+
+如遇到问题，敬请查阅 [FAQ](https://github.com/QwenLM/Qwen-VL/blob/master/FAQ_zh.md)以及issue区，如仍无法解决再提交issue。
+
+If you meet problems, please refer to [FAQ](https://github.com/QwenLM/Qwen-VL/blob/master/FAQ.md) and the issues first to search a solution before you launch a new issue.
+<br>
+
+## 使用协议 (License Agreement)
+
+研究人员与开发者可使用Qwen-VL和Qwen-VL-Chat或进行二次开发。我们同样允许商业使用，具体细节请查看[LICENSE](https://github.com/QwenLM/Qwen-VL/blob/master/LICENSE)。如需商用，请填写[问卷](https://dashscope.console.aliyun.com/openModelApply/qianwen)申请。
+
+Researchers and developers are free to use the codes and model weights of both Qwen-VL and Qwen-VL-Chat. We also allow their commercial use. Check our license at [LICENSE](LICENSE) for more details.
+<br>
+
+## 引用 (Citation)
+
+如果你觉得我们的论文和代码对你的研究有帮助，请考虑:star: 和引用 :pencil: :)
+
+If you find our paper and code useful in your research, please consider giving a star :star: and citation :pencil: :)
+
+```BibTeX
+@article{Qwen-VL,
+  title={Qwen-VL: A Frontier Large Vision-Language Model with Versatile Abilities},
+  author={Bai, Jinze and Bai, Shuai and Yang, Shusheng and Wang, Shijie and Tan, Sinan and Wang, Peng and Lin, Junyang and Zhou, Chang and Zhou, Jingren},
+  journal={arXiv preprint arXiv:2308.12966},
+  year={2023}
+}
+```
+<br>
+
+## 联系我们 (Contact Us)
+
+如果你想给我们的研发团队和产品团队留言，请通过邮件（[email protected]）联系我们。
+
+If you are interested to leave a message to either our research team or product team, feel free to send an email to [email protected].
+
+```
+
+```
+

SimSun.ttf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca4da082cd970f0c8abaa79f213ddcbc475f7b5afabcb81b385998f9ebfbb53f
+size 10499104

config.json

@@ -0,0 +1,60 @@
+{
+  "architectures": [
+    "QWenLMHeadModel"
+  ],
+  "attn_dropout_prob": 0.0,
+  "auto_map": {
+    "AutoConfig": "configuration_qwen.QWenConfig",
+    "AutoModelForCausalLM": "modeling_qwen.QWenLMHeadModel"
+  },
+  "bf16": false,
+  "emb_dropout_prob": 0.0,
+  "fp16": true,
+  "fp32": false,
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 22016,
+  "kv_channels": 128,
+  "layer_norm_epsilon": 1e-06,
+  "max_position_embeddings": 8192,
+  "model_type": "qwen",
+  "no_bias": true,
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "onnx_safe": null,
+  "quantization_config": {
+    "bits": 4,
+    "group_size": 128,
+    "damp_percent": 0.01,
+    "desc_act": false,
+    "static_groups": false,
+    "sym": true,
+    "true_sequential": true,
+    "model_name_or_path": null,
+    "model_file_base_name": "model",
+    "quant_method": "gptq"
+  },
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "scale_attn_weights": true,
+  "seq_length": 2048,
+  "tie_word_embeddings": false,
+  "tokenizer_type": "QWenTokenizer",
+  "torch_dtype": "float16",
+  "transformers_version": "4.32.0",
+  "use_cache": true,
+  "use_dynamic_ntk": true,
+  "use_flash_attn": "auto",
+  "use_logn_attn": true,
+  "visual": {
+    "heads": 16,
+    "image_size": 448,
+    "image_start_id": 151857,
+    "layers": 48,
+    "mlp_ratio": 4.9231,
+    "output_dim": 4096,
+    "patch_size": 14,
+    "width": 1664
+  },
+  "vocab_size": 151936
+}

configuration_qwen.py

@@ -0,0 +1,65 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from transformers import PretrainedConfig
+
+
+class QWenConfig(PretrainedConfig):
+    model_type = "qwen"
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+        self,
+        vocab_size=151936,
+        hidden_size=4096,
+        num_hidden_layers=32,
+        num_attention_heads=32,
+        emb_dropout_prob=0.0,
+        attn_dropout_prob=0.0,
+        layer_norm_epsilon=1e-6,
+        initializer_range=0.02,
+        max_position_embeddings=8192,
+        scale_attn_weights=True,
+        use_cache=True,
+        bf16=False,
+        fp16=False,
+        fp32=False,
+        kv_channels=128,
+        rotary_pct=1.0,
+        rotary_emb_base=10000,
+        use_dynamic_ntk=True,
+        use_logn_attn=True,
+        use_flash_attn="auto",
+        intermediate_size=22016,
+        no_bias=True,
+        tie_word_embeddings=False,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.emb_dropout_prob = emb_dropout_prob
+        self.attn_dropout_prob = attn_dropout_prob
+        self.layer_norm_epsilon = layer_norm_epsilon
+        self.initializer_range = initializer_range
+        self.scale_attn_weights = scale_attn_weights
+        self.use_cache = use_cache
+        self.max_position_embeddings = max_position_embeddings
+        self.bf16 = bf16
+        self.fp16 = fp16
+        self.fp32 = fp32
+        self.kv_channels = kv_channels
+        self.rotary_pct = rotary_pct
+        self.rotary_emb_base = rotary_emb_base
+        self.use_dynamic_ntk = use_dynamic_ntk
+        self.use_logn_attn = use_logn_attn
+        self.use_flash_attn = use_flash_attn
+        self.no_bias = no_bias
+        super().__init__(
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs
+        )

generation_config.json

@@ -0,0 +1,11 @@
+{
+  "chat_format": "chatml",
+  "eos_token_id": 151643,
+  "pad_token_id": 151643,
+  "max_window_size": 6144,
+  "max_new_tokens": 512,
+  "do_sample": true,
+  "top_k": 0,
+  "top_p": 0.5,
+  "transformers_version": "4.31.0"
+}

model-00001-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:181e991a19530b60e787c4fa8647e96fa92824934db1606e6369bbbbde88f398
+size 1982145984

model-00002-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5880b52354f16981644c2f101d9d709c0213b9863b0a8d7244f574c6c5b5f789
+size 1978254600

model-00003-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8c9b6d9bb9a825cfc48030bba6b92288d5359563e142b3f409aa4737eb59aa6
+size 1996306256

model-00004-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7380283a20d7f0adbc39d995d81767fb424c3241e78793c0e7de75756b6d43d4
+size 1994851088

model-00005-of-00005.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e5dee8e787165e11fa469fe6c08984f731647e72265b781f12e08c60d51893b
+size 1780398232

modeling_qwen.py

@@ -0,0 +1,1154 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import importlib
+import math
+from typing import TYPE_CHECKING, Optional, Tuple, Union, Callable, List, Any, Generator
+
+import torch
+import torch.nn.functional as F
+import torch.utils.checkpoint
+from torch.cuda.amp import autocast
+
+from torch.nn import CrossEntropyLoss
+from transformers import PreTrainedTokenizer, GenerationConfig, StoppingCriteriaList
+from transformers.generation.logits_process import LogitsProcessorList
+
+if TYPE_CHECKING:
+    from transformers.generation.streamers import BaseStreamer
+from transformers.generation.utils import GenerateOutput
+from transformers.modeling_outputs import (
+    BaseModelOutputWithPast,
+    CausalLMOutputWithPast,
+)
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import logging
+
+try:
+    from einops import rearrange
+except ImportError:
+    rearrange = None
+from torch import nn
+
+SUPPORT_CUDA = torch.cuda.is_available()
+SUPPORT_BF16 = SUPPORT_CUDA and torch.cuda.is_bf16_supported()
+SUPPORT_FP16 = SUPPORT_CUDA and torch.cuda.get_device_capability(0)[0] >= 7
+
+from .configuration_qwen import QWenConfig
+from .qwen_generation_utils import (
+    HistoryType,
+    make_context,
+    decode_tokens,
+    get_stop_words_ids,
+    StopWordsLogitsProcessor,
+)
+from .visual import VisionTransformer
+
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "qwen"
+_CONFIG_FOR_DOC = "QWenConfig"
+
+QWen_PRETRAINED_MODEL_ARCHIVE_LIST = ["qwen-7b"]
+
+_ERROR_BAD_CHAT_FORMAT = """\
+We detect you are probably using the pretrained model (rather than chat model) for chatting, since the chat_format in generation_config is not "chatml".
+If you are directly using the model downloaded from Huggingface, please make sure you are using our "Qwen/Qwen-7B-Chat" Huggingface model (rather than "Qwen/Qwen-7B") when you call model.chat().
+我们检测到您可能在使用预训练模型（而非chat模型）进行多轮chat，因为您当前在generation_config指定的chat_format，并未设置为我们在对话中所支持的"chatml"格式。
+如果您在直接使用我们从Huggingface提供的模型，请确保您在调用model.chat()时，使用的是"Qwen/Qwen-7B-Chat"模型（而非"Qwen/Qwen-7B"预训练模型）。
+"""
+
+_SENTINEL = object()
+_ERROR_STREAM_IN_CHAT = """\
+Pass argument `stream` to model.chat() is buggy, deprecated, and marked for removal. Please use model.chat_stream(...) instead of model.chat(..., stream=True).
+向model.chat()传入参数stream的用法可能存在Bug，该用法已被废弃，将在未来被移除。请使用model.chat_stream(...)代替model.chat(..., stream=True)。
+"""
+
+apply_rotary_emb_func = None
+rms_norm = None
+
+
+# Copied from transformers.models.bart.modeling_bart._make_causal_mask
+def _make_causal_mask(
+    input_ids_shape: torch.Size, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0
+):
+    """
+    Make causal mask used for bi-directional self-attention.
+    """
+    bsz, tgt_len = input_ids_shape
+    mask = torch.full((tgt_len, tgt_len), torch.finfo(dtype).min, device=device)
+    mask_cond = torch.arange(mask.size(-1), device=device)
+    mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
+    mask = mask.to(dtype)
+
+    if past_key_values_length > 0:
+        mask = torch.cat([torch.zeros(tgt_len, past_key_values_length, dtype=dtype, device=device), mask], dim=-1)
+    return mask[None, None, :, :].expand(bsz, 1, tgt_len, tgt_len + past_key_values_length)
+
+
+# Copied from transformers.models.bart.modeling_bart._expand_mask
+def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None):
+    """
+    Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.
+    """
+    bsz, src_len = mask.size()
+    tgt_len = tgt_len if tgt_len is not None else src_len
+
+    expanded_mask = mask[:, None, None, :].expand(bsz, 1, tgt_len, src_len).to(dtype)
+
+    inverted_mask = 1.0 - expanded_mask
+
+    return inverted_mask.masked_fill(inverted_mask.to(torch.bool), torch.finfo(dtype).min)
+
+
+class QWenAttention(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+
+        self.register_buffer("masked_bias", torch.tensor(-1e4), persistent=False)
+        self.seq_length = config.seq_length
+
+        self.hidden_size = config.hidden_size
+        self.split_size = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.head_dim = self.hidden_size // self.num_heads
+
+        self.scale_attn_weights = True
+
+        self.projection_size = config.kv_channels * config.num_attention_heads
+
+        assert self.projection_size % config.num_attention_heads == 0
+        self.hidden_size_per_attention_head = (
+            self.projection_size // config.num_attention_heads
+        )
+
+        self.c_attn = nn.Linear(config.hidden_size, 3 * self.projection_size)
+
+        self.c_proj = nn.Linear(
+            config.hidden_size, self.projection_size, bias=not config.no_bias
+        )
+
+        self.is_fp32 = not (config.bf16 or config.fp16)
+        self.bf16 = config.bf16
+
+        self.use_dynamic_ntk = config.use_dynamic_ntk
+        self.use_logn_attn = config.use_logn_attn
+
+        logn_list = [
+            math.log(i, self.seq_length) if i > self.seq_length else 1
+            for i in range(1, 32768)
+        ]
+        self.logn_tensor = torch.tensor(logn_list)[None, :, None, None]
+
+        self.attn_dropout = nn.Dropout(config.attn_dropout_prob)
+
+    def _attn(self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None):
+        attn_weights = torch.matmul(query, key.transpose(-1, -2))
+
+        if self.scale_attn_weights:
+            attn_weights = attn_weights / torch.full(
+                [],
+                value.size(-1) ** 0.5,
+                dtype=attn_weights.dtype,
+                device=attn_weights.device,
+            )
+
+        query_length, key_length = query.size(-2), key.size(-2)
+        # causal_mask = self.bias[
+        #     :, :, key_length - query_length : key_length, :key_length
+        # ]
+        # mask_value = torch.finfo(attn_weights.dtype).min
+        # mask_value = torch.full([], mask_value, dtype=attn_weights.dtype).to(
+        #     attn_weights.device
+        # )
+        # attn_weights = torch.where(
+        #     causal_mask, attn_weights.to(attn_weights.dtype), mask_value
+        # )
+        attn_weights = attn_weights + attention_mask
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+
+        attn_weights = attn_weights.type(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+        attn_output = attn_output.transpose(1, 2)
+
+        return attn_output, attn_weights
+
+    def _upcast_and_reordered_attn(
+        self, query, key, value, registered_causal_mask, attention_mask=None, head_mask=None
+    ):
+        bsz, num_heads, q_seq_len, dk = query.size()
+        _, _, k_seq_len, _ = key.size()
+
+        attn_weights = torch.empty(
+            bsz * num_heads,
+            q_seq_len,
+            k_seq_len,
+            dtype=torch.float32,
+            device=query.device,
+        )
+
+        scale_factor = 1.0
+        if self.scale_attn_weights:
+            scale_factor /= float(value.size(-1)) ** 0.5
+
+        with autocast(enabled=False):
+            q, k = query.reshape(-1, q_seq_len, dk), key.transpose(-1, -2).reshape(
+                -1, dk, k_seq_len
+            )
+            attn_weights = torch.baddbmm(
+                attn_weights, q.float(), k.float(), beta=0, alpha=scale_factor
+            )
+            attn_weights = attn_weights.reshape(bsz, num_heads, q_seq_len, k_seq_len)
+
+        query_length, key_length = query.size(-2), key.size(-2)
+        causal_mask = registered_causal_mask[
+            :, :, key_length - query_length : key_length, :key_length
+        ]
+        mask_value = torch.finfo(attn_weights.dtype).min
+        mask_value = torch.tensor(mask_value, dtype=attn_weights.dtype).to(
+            attn_weights.device
+        )
+        attn_weights = torch.where(causal_mask, attn_weights, mask_value)
+
+        if attention_mask is not None:
+            attn_weights = attn_weights + attention_mask
+
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1)
+
+        if attn_weights.dtype != torch.float32:
+            raise RuntimeError(
+                "Error with upcasting, attn_weights does not have dtype torch.float32"
+            )
+        attn_weights = attn_weights.type(value.dtype)
+        attn_weights = self.attn_dropout(attn_weights)
+
+        if head_mask is not None:
+            attn_weights = attn_weights * head_mask
+
+        attn_output = torch.matmul(attn_weights, value)
+
+        return attn_output, attn_weights
+
+    def _split_heads(self, tensor, num_heads, attn_head_size):
+        new_shape = tensor.size()[:-1] + (num_heads, attn_head_size)
+        tensor = tensor.view(new_shape)
+        return tensor
+
+    def _merge_heads(self, tensor, num_heads, attn_head_size):
+        tensor = tensor.contiguous()
+        new_shape = tensor.size()[:-2] + (num_heads * attn_head_size,)
+        return tensor.view(new_shape)
+
+    def forward(
+        self,
+        hidden_states: Optional[Tuple[torch.FloatTensor]],
+        rotary_pos_emb: Optional[List[torch.Tensor]] = None,
+        registered_causal_mask: Optional[torch.Tensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+    ):
+
+        mixed_x_layer = self.c_attn(hidden_states)
+
+        query, key, value = mixed_x_layer.split(self.split_size, dim=2)
+
+        query = self._split_heads(query, self.num_heads, self.head_dim)
+        key = self._split_heads(key, self.num_heads, self.head_dim)
+        value = self._split_heads(value, self.num_heads, self.head_dim)
+
+        if rotary_pos_emb is not None:
+            cur_len = query.shape[1]
+            rotary_pos_emb = [i[:, -cur_len:, :, :] for i in rotary_pos_emb]
+            rotary_pos_emb = (rotary_pos_emb,) * 2
+            q_pos_emb, k_pos_emb = rotary_pos_emb
+            # Slice the pos emb for current inference
+            query = apply_rotary_pos_emb(query, q_pos_emb)
+            key = apply_rotary_pos_emb(key, k_pos_emb)
+
+        if layer_past is not None:
+            past_key, past_value = layer_past[0], layer_past[1]
+            key = torch.cat((past_key, key), dim=1)
+            value = torch.cat((past_value, value), dim=1)
+
+        if use_cache:
+            present = (key, value)
+        else:
+            present = None
+
+        if self.use_logn_attn and not self.training:
+            if self.logn_tensor.device != query.device or self.logn_tensor.dtype != query.dtype:
+                self.logn_tensor = self.logn_tensor.to(query.device).type_as(query)
+            seq_start = key.size(1) - query.size(1)
+            seq_end = key.size(1)
+            logn_tensor = self.logn_tensor[:, seq_start:seq_end, :, :]
+            query = query * logn_tensor.expand_as(query)
+
+        query = query.permute(0, 2, 1, 3)
+        key = key.permute(0, 2, 1, 3)
+        value = value.permute(0, 2, 1, 3)
+        attn_output, attn_weight = self._attn(
+            query, key, value, registered_causal_mask, attention_mask, head_mask
+        )
+        context_layer = self._merge_heads(
+            attn_output, self.num_heads, self.head_dim
+        )
+
+        attn_output = self.c_proj(context_layer)
+
+        outputs = (attn_output, present)
+        if output_attentions:
+            outputs += (attn_weight,)
+
+        return outputs
+
+
+class QWenMLP(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.w1 = nn.Linear(
+            config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias
+        )
+        self.w2 = nn.Linear(
+            config.hidden_size, config.intermediate_size // 2, bias=not config.no_bias
+        )
+        ff_dim_in = config.intermediate_size // 2
+        self.c_proj = nn.Linear(ff_dim_in, config.hidden_size, bias=not config.no_bias)
+
+    def forward(self, hidden_states):
+        a1 = self.w1(hidden_states)
+        a2 = self.w2(hidden_states)
+        intermediate_parallel = a1 * F.silu(a2)
+        output = self.c_proj(intermediate_parallel)
+        return output
+
+class QWenBlock(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        hidden_size = config.hidden_size
+        self.bf16 = config.bf16
+
+        self.ln_1 = RMSNorm(
+            hidden_size,
+            eps=config.layer_norm_epsilon,
+        )
+        self.attn = QWenAttention(config)
+        self.ln_2 = RMSNorm(
+            hidden_size,
+            eps=config.layer_norm_epsilon,
+        )
+
+        self.mlp = QWenMLP(config)
+
+    def forward(
+        self,
+        hidden_states: Optional[Tuple[torch.FloatTensor]],
+        rotary_pos_emb: Optional[List[torch.Tensor]] = None,
+        registered_causal_mask: Optional[torch.Tensor] = None,
+        layer_past: Optional[Tuple[torch.Tensor]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = False,
+        output_attentions: Optional[bool] = False,
+    ):
+        layernorm_output = self.ln_1(hidden_states)
+
+        attn_outputs = self.attn(
+            layernorm_output,
+            rotary_pos_emb,
+            registered_causal_mask=registered_causal_mask,
+            layer_past=layer_past,
+            attention_mask=attention_mask,
+            head_mask=head_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+        )
+        attn_output = attn_outputs[0]
+
+        outputs = attn_outputs[1:]
+
+        residual = hidden_states
+        layernorm_input = attn_output + residual
+
+        layernorm_output = self.ln_2(layernorm_input)
+
+        residual = layernorm_input
+        mlp_output = self.mlp(layernorm_output)
+        hidden_states = residual + mlp_output
+
+        if use_cache:
+            outputs = (hidden_states,) + outputs
+        else:
+            outputs = (hidden_states,) + outputs[1:]
+
+        return outputs
+
+
+class QWenPreTrainedModel(PreTrainedModel):
+    config_class = QWenConfig
+    base_model_prefix = "transformer"
+    is_parallelizable = False
+    supports_gradient_checkpointing = True
+    _no_split_modules = ["QWenBlock"]
+
+    def __init__(self, *inputs, **kwargs):
+        super().__init__(*inputs, **kwargs)
+
+    def _init_weights(self, module):
+        """Initialize the weights."""
+        if isinstance(module, nn.Linear):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.bias is not None:
+                module.bias.data.zero_()
+        elif isinstance(module, nn.Embedding):
+            module.weight.data.normal_(mean=0.0, std=self.config.initializer_range)
+            if module.padding_idx is not None:
+                module.weight.data[module.padding_idx].zero_()
+        elif isinstance(module, RMSNorm):
+            module.weight.data.fill_(1.0)
+
+        for name, p in module.named_parameters():
+            if name == "c_proj.weight":
+                p.data.normal_(
+                    mean=0.0,
+                    std=(
+                        self.config.initializer_range
+                        / math.sqrt(2 * self.config.num_hidden_layers)
+                    ),
+                )
+
+    def _set_gradient_checkpointing(self, module, value=False):
+        if isinstance(module, QWenModel):
+            module.gradient_checkpointing = value
+
+
+class QWenModel(QWenPreTrainedModel):
+    _keys_to_ignore_on_load_missing = ["attn.masked_bias"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.vocab_size = config.vocab_size
+        self.num_hidden_layers = config.num_hidden_layers
+        self.embed_dim = config.hidden_size
+
+        self.gradient_checkpointing = False
+        self.use_dynamic_ntk = config.use_dynamic_ntk
+        self.seq_length = config.seq_length
+
+        self.wte = nn.Embedding(self.vocab_size, self.embed_dim)
+
+        self.drop = nn.Dropout(config.emb_dropout_prob)
+
+        if config.rotary_pct == 1.0:
+            self.rotary_ndims = None
+        else:
+            assert config.rotary_pct < 1
+            self.rotary_ndims = int(
+                config.kv_channels * config.rotary_pct
+            )
+        dim = (
+            self.rotary_ndims
+            if self.rotary_ndims is not None
+            else config.kv_channels
+        )
+        self.rotary_emb = RotaryEmbedding(dim, base=config.rotary_emb_base)
+
+        self.use_flash_attn = config.use_flash_attn
+        self.is_fp32 = not (config.bf16 or config.fp16)
+        self.registered_causal_mask = None
+        # if (
+        #     self.use_flash_attn
+        #     and flash_attn_unpadded_func is not None
+        #     and not self.is_fp32
+        # ):
+        #     self.registered_causal_mask = None
+        # else:
+        #     max_positions = config.max_position_embeddings
+        #     self.register_buffer(
+        #         "registered_causal_mask",
+        #         torch.tril(
+        #             torch.ones((max_positions, max_positions), dtype=torch.bool)
+        #         ).view(1, 1, max_positions, max_positions),
+        #         persistent=False,
+        #     )
+
+        self.h = nn.ModuleList(
+            [
+                QWenBlock(
+                    config
+                )
+                for i in range(config.num_hidden_layers)
+            ]
+        )
+        self.ln_f = RMSNorm(
+            self.embed_dim,
+            eps=config.layer_norm_epsilon,
+        )
+
+        self.visual = VisionTransformer(**config.visual)
+
+        self.post_init()
+
+    def get_input_embeddings(self):
+        return self.wte
+
+    def set_input_embeddings(self, new_embeddings):
+        self.wte = new_embeddings
+    
+    # Copied from transformers.models.bart.modeling_bart.BartDecoder._prepare_decoder_attention_mask
+    def _prepare_decoder_attention_mask(self, attention_mask, input_shape, inputs_embeds, past_key_values_length):
+        # create causal mask
+        # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+        combined_attention_mask = None
+        if input_shape[-1] > 1:
+            combined_attention_mask = _make_causal_mask(
+                input_shape,
+                inputs_embeds.dtype,
+                device=inputs_embeds.device,
+                past_key_values_length=past_key_values_length,
+            )
+
+        if attention_mask is not None:
+            # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
+            expanded_attn_mask = _expand_mask(attention_mask, inputs_embeds.dtype, tgt_len=input_shape[-1]).to(
+                inputs_embeds.device
+            )
+            combined_attention_mask = (
+                expanded_attn_mask if combined_attention_mask is None else expanded_attn_mask + combined_attention_mask
+            )
+
+        return combined_attention_mask
+
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        token_type_ids: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ):
+        if past_key_values is None and torch.any(input_ids == self.config.visual['image_start_id']):
+            bos_pos = torch.where(input_ids == self.config.visual['image_start_id'])
+            eos_pos = torch.where(input_ids == self.config.visual['image_start_id'] + 1)
+            assert (bos_pos[0] == eos_pos[0]).all()
+            img_pos = torch.stack((bos_pos[0], bos_pos[1], eos_pos[1]), dim=1)
+            images = []
+            for i, a, b in img_pos:
+                image = input_ids[i][a + 1 : b - 1].tolist()
+                image = image[ : image.index(self.config.visual['image_start_id'] + 2)]
+                images.append(bytes(image).decode('utf-8'))
+
+            images = self.visual.encode(images)
+            assert images.shape[0] == len(images)
+        else:
+            images = None
+
+        output_attentions = (
+            output_attentions
+            if output_attentions is not None
+            else self.config.output_attentions
+        )
+        output_hidden_states = (
+            output_hidden_states
+            if output_hidden_states is not None
+            else self.config.output_hidden_states
+        )
+        use_cache = use_cache if use_cache is not None else self.config.use_cache
+        return_dict = (
+            return_dict if return_dict is not None else self.config.use_return_dict
+        )
+
+        if input_ids is not None and inputs_embeds is not None:
+            raise ValueError(
+                "You cannot specify both input_ids and inputs_embeds at the same time"
+            )
+        elif input_ids is not None:
+            input_shape = input_ids.size()
+            input_ids = input_ids.view(-1, input_shape[-1])
+            batch_size = input_ids.shape[0]
+        elif inputs_embeds is not None:
+            input_shape = inputs_embeds.size()[:-1]
+            batch_size = inputs_embeds.shape[0]
+        else:
+            raise ValueError("You have to specify either input_ids or inputs_embeds")
+
+        device = input_ids.device if input_ids is not None else inputs_embeds.device
+
+        if token_type_ids is not None:
+            token_type_ids = token_type_ids.view(-1, input_shape[-1])
+        if position_ids is not None:
+            position_ids = position_ids.view(-1, input_shape[-1])
+
+        if past_key_values is None:
+            past_length = 0
+            past_key_values = tuple([None] * len(self.h))
+        else:
+            past_length = past_key_values[0][0].size(-2)
+
+        if position_ids is None:
+            position_ids = torch.arange(
+                past_length,
+                input_shape[-1] + past_length,
+                dtype=torch.long,
+                device=device,
+            )
+            position_ids = position_ids.unsqueeze(0).view(-1, input_shape[-1])
+
+        encoder_attention_mask = None
+        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)
+
+        if inputs_embeds is None:
+            inputs_embeds = self.wte(input_ids)
+
+        if batch_size <= 0:
+            raise ValueError("batch_size has to be defined and > 0")
+        attention_mask = self._prepare_decoder_attention_mask(
+            attention_mask, input_shape, inputs_embeds, past_length
+        )
+
+        hidden_states = inputs_embeds
+
+        kv_seq_len = hidden_states.size()[1]
+        if past_key_values[0] is not None:
+            # past key values[0][0] shape: bs * seq_len * head_num * dim
+            kv_seq_len += past_key_values[0][0].shape[1]
+        if (
+            self.use_dynamic_ntk
+            and kv_seq_len == hidden_states.size()[1]
+            and not self.training
+        ):
+            context_value = math.log(kv_seq_len / self.seq_length, 2) + 1
+            ntk_alpha = 2 ** math.ceil(context_value) - 1
+            ntk_alpha = max(ntk_alpha, 1)
+        else:
+            ntk_alpha = self.rotary_emb._ntk_alpha_cached
+
+        rotary_pos_emb = self.rotary_emb(kv_seq_len, ntk_alpha=ntk_alpha)
+        for idx in range(len(rotary_pos_emb)):
+            rotary_pos_emb[idx] = rotary_pos_emb[idx].to(hidden_states.device)
+
+        hidden_states = self.drop(hidden_states)
+        if images is not None:
+            for idx, (i, a, b) in enumerate(img_pos):
+                hidden_states[i][a + 1 : b] = images[idx]
+        output_shape = input_shape + (hidden_states.size(-1),)
+
+        if self.gradient_checkpointing and self.training:
+            if use_cache:
+                logger.warning_once(
+                    "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                )
+                use_cache = False
+
+        presents = () if use_cache else None
+        all_self_attentions = () if output_attentions else None
+        all_hidden_states = () if output_hidden_states else None
+        for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
+
+            if output_hidden_states:
+                all_hidden_states = all_hidden_states + (hidden_states,)
+
+            if self.gradient_checkpointing and self.training:
+
+                def create_custom_forward(module):
+                    def custom_forward(*inputs):
+                        # None for past_key_value
+                        return module(*inputs, use_cache, output_attentions)
+
+                    return custom_forward
+
+                outputs = torch.utils.checkpoint.checkpoint(
+                    create_custom_forward(block),
+                    hidden_states,
+                    rotary_pos_emb,
+                    self.registered_causal_mask,
+                    None,
+                    attention_mask,
+                    head_mask[i],
+                    encoder_hidden_states,
+                    encoder_attention_mask,
+                )
+            else:
+                outputs = block(
+                    hidden_states,
+                    layer_past=layer_past,
+                    rotary_pos_emb=rotary_pos_emb,
+                    registered_causal_mask=self.registered_causal_mask,
+                    attention_mask=attention_mask,
+                    head_mask=head_mask[i],
+                    encoder_hidden_states=encoder_hidden_states,
+                    encoder_attention_mask=encoder_attention_mask,
+                    use_cache=use_cache,
+                    output_attentions=output_attentions,
+                )
+
+            hidden_states = outputs[0]
+            if use_cache is True:
+                presents = presents + (outputs[1],)
+
+            if output_attentions:
+                all_self_attentions = all_self_attentions + (outputs[2 if use_cache else 1],)
+
+        hidden_states = self.ln_f(hidden_states)
+        hidden_states = hidden_states.view(output_shape)
+        # Add last hidden state
+        if output_hidden_states:
+            all_hidden_states = all_hidden_states + (hidden_states,)
+
+        if not return_dict:
+            return tuple(
+                v for v in [hidden_states, presents, all_hidden_states] if v is not None
+            )
+
+        return BaseModelOutputWithPast(
+            last_hidden_state=hidden_states,
+            past_key_values=presents,
+            hidden_states=all_hidden_states,
+            attentions=all_self_attentions,
+        )
+
+
+class QWenLMHeadModel(QWenPreTrainedModel):
+    _keys_to_ignore_on_load_missing = [r"h\.\d+\.attn\.rotary_emb\.inv_freq"]
+    _keys_to_ignore_on_load_unexpected = [r"h\.\d+\.attn\.masked_bias"]
+
+    def __init__(self, config):
+        super().__init__(config)
+        assert (
+            config.bf16 + config.fp16 + config.fp32 <= 1
+        ), "Only one of \"bf16\", \"fp16\", \"fp32\" can be true"
+
+        autoset_precision = config.bf16 + config.fp16 + config.fp32 == 0
+
+        if autoset_precision:
+            if SUPPORT_BF16:
+                logger.warn(
+                    "The model is automatically converting to bf16 for faster inference. "
+                    "If you want to disable the automatic precision, please manually add bf16/fp16/fp32=True to \"AutoModelForCausalLM.from_pretrained\"."
+                )
+                config.bf16 = True
+            elif SUPPORT_FP16:
+                logger.warn(
+                    "The model is automatically converting to fp16 for faster inference. "
+                    "If you want to disable the automatic precision, please manually add bf16/fp16/fp32=True to \"AutoModelForCausalLM.from_pretrained\"."
+                )
+                config.fp16 = True
+            else:
+                config.fp32 = True
+
+        if config.bf16 and SUPPORT_CUDA and not SUPPORT_BF16:
+            logger.warn("Your device does NOT seem to support bf16, you can switch to fp16 or fp32 by by passing fp16/fp32=True in \"AutoModelForCausalLM.from_pretrained\".")
+        if config.fp16 and SUPPORT_CUDA and not SUPPORT_FP16:
+            logger.warn("Your device does NOT support faster inference with fp16, please switch to fp32 which is likely to be faster")
+        if config.fp32:
+            if SUPPORT_BF16:
+                logger.warn("Your device support faster inference by passing bf16=True in \"AutoModelForCausalLM.from_pretrained\".")
+            elif SUPPORT_FP16:
+                logger.warn("Your device support faster inference by passing fp16=True in \"AutoModelForCausalLM.from_pretrained\".")
+
+        self.transformer = QWenModel(config)
+        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
+
+        if config.bf16:
+            self.transformer.bfloat16()
+            self.lm_head.bfloat16()
+        if config.fp16:
+            self.transformer.half()
+            self.lm_head.half()
+        self.post_init()
+
+    def get_output_embeddings(self):
+        return self.lm_head
+
+    def set_output_embeddings(self, new_embeddings):
+        self.lm_head = new_embeddings
+
+    def prepare_inputs_for_generation(
+        self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs
+    ):
+        token_type_ids = kwargs.get("token_type_ids", None)
+        if past_key_values:
+            input_ids = input_ids[:, -1].unsqueeze(-1)
+            if token_type_ids is not None:
+                token_type_ids = token_type_ids[:, -1].unsqueeze(-1)
+
+        attention_mask = kwargs.get("attention_mask", None)
+        position_ids = kwargs.get("position_ids", None)
+
+        if attention_mask is not None and position_ids is None:
+            position_ids = attention_mask.long().cumsum(-1) - 1
+            position_ids.masked_fill_(attention_mask == 0, 1)
+            if past_key_values:
+                position_ids = position_ids[:, -1].unsqueeze(-1)
+        else:
+            position_ids = None
+
+        if inputs_embeds is not None and past_key_values is None:
+            model_inputs = {"inputs_embeds": inputs_embeds}
+        else:
+            model_inputs = {"input_ids": input_ids}
+
+        model_inputs.update(
+            {
+                "past_key_values": past_key_values,
+                "use_cache": kwargs.get("use_cache"),
+                "position_ids": position_ids,
+                "attention_mask": attention_mask,
+                "token_type_ids": token_type_ids,
+            }
+        )
+        return model_inputs
+
+    def forward(
+        self,
+        input_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
+        attention_mask: Optional[torch.FloatTensor] = None,
+        token_type_ids: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        head_mask: Optional[torch.FloatTensor] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        encoder_hidden_states: Optional[torch.Tensor] = None,
+        encoder_attention_mask: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+
+        return_dict = (
+            return_dict if return_dict is not None else self.config.use_return_dict
+        )
+
+        transformer_outputs = self.transformer(
+            input_ids,
+            past_key_values=past_key_values,
+            attention_mask=attention_mask,
+            token_type_ids=token_type_ids,
+            position_ids=position_ids,
+            head_mask=head_mask,
+            inputs_embeds=inputs_embeds,
+            encoder_hidden_states=encoder_hidden_states,
+            encoder_attention_mask=encoder_attention_mask,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+        )
+        hidden_states = transformer_outputs[0]
+
+        lm_logits = self.lm_head(hidden_states)
+
+        loss = None
+        if labels is not None:
+            labels = labels.to(lm_logits.device)
+            shift_logits = lm_logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            loss_fct = CrossEntropyLoss()
+            loss = loss_fct(
+                shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1)
+            )
+
+        if not return_dict:
+            output = (lm_logits,) + transformer_outputs[1:]
+            return ((loss,) + output) if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=lm_logits,
+            past_key_values=transformer_outputs.past_key_values,
+            hidden_states=transformer_outputs.hidden_states,
+            attentions=transformer_outputs.attentions,
+        )
+
+    @staticmethod
+    def _reorder_cache(
+        past_key_values: Tuple[Tuple[torch.Tensor]], beam_idx: torch.Tensor
+    ) -> Tuple[Tuple[torch.Tensor]]:
+
+        return tuple(
+            tuple(
+                past_state.index_select(0, beam_idx.to(past_state.device))
+                for past_state in layer_past
+            )
+            for layer_past in past_key_values
+        )
+
+    def chat(
+        self,
+        tokenizer: PreTrainedTokenizer,
+        query: str,
+        history: Optional[HistoryType],
+        system: str = "You are a helpful assistant.",
+        append_history: bool = True,
+        stream: Optional[bool] = _SENTINEL,
+        stop_words_ids: Optional[List[List[int]]] = None,
+        generation_config: Optional[GenerationConfig] = None,
+        **kwargs,
+    ) -> Tuple[str, HistoryType]:
+        generation_config = generation_config if generation_config is not None else self.generation_config
+
+        assert stream is _SENTINEL, _ERROR_STREAM_IN_CHAT
+        assert generation_config.chat_format == 'chatml', _ERROR_BAD_CHAT_FORMAT
+        if history is None:
+            history = []
+        if stop_words_ids is None:
+            stop_words_ids = []
+
+        max_window_size = kwargs.get('max_window_size', None)
+        if max_window_size is None:
+            max_window_size = generation_config.max_window_size
+        raw_text, context_tokens = make_context(
+            tokenizer,
+            query,
+            history=history,
+            system=system,
+            max_window_size=max_window_size,
+            chat_format=generation_config.chat_format,
+        )
+
+        stop_words_ids.extend(get_stop_words_ids(
+            generation_config.chat_format, tokenizer
+        ))
+        input_ids = torch.tensor([context_tokens]).to(self.device)
+        outputs = self.generate(
+                    input_ids,
+                    stop_words_ids=stop_words_ids,
+                    return_dict_in_generate=False,
+                    generation_config=generation_config,
+                    **kwargs,
+                )
+
+        response = decode_tokens(
+            outputs[0],
+            tokenizer,
+            raw_text_len=len(raw_text),
+            context_length=len(context_tokens),
+            chat_format=generation_config.chat_format,
+            verbose=False,
+            errors='replace'
+        )
+
+        if append_history:
+            history.append((query, response))
+
+        return response, history
+
+    def chat_stream(
+            self,
+            tokenizer: PreTrainedTokenizer,
+            query: str,
+            history: Optional[HistoryType],
+            system: str = "You are a helpful assistant.",
+            stop_words_ids: Optional[List[List[int]]] = None,
+            logits_processor: Optional[LogitsProcessorList] = None,
+            generation_config: Optional[GenerationConfig] = None,
+            **kwargs,
+    ) -> Generator[str, Any, None]:
+        generation_config = generation_config if generation_config is not None else self.generation_config
+        assert generation_config.chat_format == 'chatml', _ERROR_BAD_CHAT_FORMAT
+        if history is None:
+            history = []
+        if stop_words_ids is None:
+            stop_words_ids = []
+
+        max_window_size = kwargs.get('max_window_size', None)
+        if max_window_size is None:
+            max_window_size = generation_config.max_window_size
+        raw_text, context_tokens = make_context(
+            tokenizer,
+            query,
+            history=history,
+            system=system,
+            max_window_size=max_window_size,
+            chat_format=generation_config.chat_format,
+        )
+
+        stop_words_ids.extend(get_stop_words_ids(
+            generation_config.chat_format, tokenizer
+        ))
+        if stop_words_ids is not None:
+            stop_words_logits_processor = StopWordsLogitsProcessor(
+                stop_words_ids=stop_words_ids,
+                eos_token_id=generation_config.eos_token_id,
+            )
+            if logits_processor is None:
+                logits_processor = LogitsProcessorList([stop_words_logits_processor])
+            else:
+                logits_processor.append(stop_words_logits_processor)
+        input_ids = torch.tensor([context_tokens]).to(self.device)
+
+        from transformers_stream_generator.main import NewGenerationMixin, StreamGenerationConfig
+        self.__class__.generate_stream = NewGenerationMixin.generate
+        self.__class__.sample_stream = NewGenerationMixin.sample_stream
+        stream_config = StreamGenerationConfig(**generation_config.to_dict(), do_stream=True)
+
+        def stream_generator():
+            outputs = []
+            for token in self.generate_stream(
+                    input_ids,
+                    return_dict_in_generate=False,
+                    generation_config=stream_config,
+                    logits_processor=logits_processor,
+                    seed=-1,
+                    **kwargs):
+                outputs.append(token.item())
+                yield tokenizer.decode(outputs, skip_special_tokens=True, errors='ignore')
+
+        return stream_generator()
+
+    def generate(
+        self,
+        inputs: Optional[torch.Tensor] = None,
+        generation_config: Optional[GenerationConfig] = None,
+        logits_processor: Optional[LogitsProcessorList] = None,
+        stopping_criteria: Optional[StoppingCriteriaList] = None,
+        prefix_allowed_tokens_fn: Optional[
+            Callable[[int, torch.Tensor], List[int]]
+        ] = None,
+        synced_gpus: Optional[bool] = None,
+        assistant_model: Optional["PreTrainedModel"] = None,
+        streamer: Optional["BaseStreamer"] = None,
+        **kwargs,
+    ) -> Union[GenerateOutput, torch.LongTensor]:
+        generation_config = generation_config if generation_config is not None else self.generation_config
+
+        # Process stop_words_ids.
+        stop_words_ids = kwargs.pop("stop_words_ids", None)
+        if stop_words_ids is None and generation_config is not None:
+            stop_words_ids = getattr(generation_config, "stop_words_ids", None)
+        if stop_words_ids is None:
+            stop_words_ids = getattr(generation_config, "stop_words_ids", None)
+
+        if stop_words_ids is not None:
+            stop_words_logits_processor = StopWordsLogitsProcessor(
+                stop_words_ids=stop_words_ids,
+                eos_token_id=generation_config.eos_token_id,
+            )
+            if logits_processor is None:
+                logits_processor = LogitsProcessorList([stop_words_logits_processor])
+            else:
+                logits_processor.append(stop_words_logits_processor)
+
+        return super().generate(
+            inputs,
+            generation_config=generation_config,
+            logits_processor=logits_processor,
+            stopping_criteria=stopping_criteria,
+            prefix_allowed_tokens_fn=prefix_allowed_tokens_fn,
+            synced_gpus=synced_gpus,
+            assistant_model=assistant_model,
+            streamer=streamer,
+            **kwargs,
+        )
+
+
+class RotaryEmbedding(torch.nn.Module):
+    def __init__(self, dim, base=10000):
+        super().__init__()
+        self.dim = dim
+        self.base = base
+        self.inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float() / dim))
+        if importlib.util.find_spec("einops") is None:
+            raise RuntimeError("einops is required for Rotary Embedding")
+
+        self._rotary_pos_emb_cache = None
+        self._seq_len_cached = 0
+        self._ntk_alpha_cached = 1.0
+
+    def update_rotary_pos_emb_cache(self, max_seq_len, offset=0, ntk_alpha=1.0):
+        seqlen = max_seq_len + offset
+        if seqlen > self._seq_len_cached or ntk_alpha != self._ntk_alpha_cached:
+            base = self.base * ntk_alpha ** (self.dim / (self.dim - 2))
+            self.inv_freq = 1.0 / (
+                base
+                ** (
+                    torch.arange(0, self.dim, 2, device=self.inv_freq.device).float()
+                    / self.dim
+                )
+            )
+            self._seq_len_cached = max(2 * seqlen, 16)
+            self._ntk_alpha_cached = ntk_alpha
+            seq = torch.arange(self._seq_len_cached, device=self.inv_freq.device)
+            freqs = torch.outer(seq.type_as(self.inv_freq), self.inv_freq)
+            
+            emb = torch.cat((freqs, freqs), dim=-1)
+            from einops import rearrange
+
+            emb = rearrange(emb, "n d -> 1 n 1 d")
+
+            cos, sin = emb.cos(), emb.sin()
+            self._rotary_pos_emb_cache = [cos, sin]
+
+    def forward(self, max_seq_len, offset=0, ntk_alpha=1.0):
+        self.update_rotary_pos_emb_cache(max_seq_len, offset, ntk_alpha)
+        cos, sin = self._rotary_pos_emb_cache
+        return [cos[:, offset : offset + max_seq_len], sin[:, offset : offset + max_seq_len]]
+
+
+def _rotate_half(x):
+    from einops import rearrange
+
+    x = rearrange(x, "... (j d) -> ... j d", j=2)
+    x1, x2 = x.unbind(dim=-2)
+    return torch.cat((-x2, x1), dim=-1)
+
+
+def apply_rotary_pos_emb(t, freqs):
+    cos, sin = freqs
+    if apply_rotary_emb_func is not None and t.is_cuda:
+        t_ = t.float()
+        cos = cos.squeeze(0).squeeze(1)[:, : cos.shape[-1] // 2]
+        sin = sin.squeeze(0).squeeze(1)[:, : sin.shape[-1] // 2]
+        output = apply_rotary_emb_func(t_, cos, sin).type_as(t)
+        return output
+    else:
+        rot_dim = freqs[0].shape[-1]
+        cos, sin = freqs
+        t_, t_pass_ = t[..., :rot_dim], t[..., rot_dim:]
+        t_ = t_.float()
+        t_pass_ = t_pass_.float()
+        t_ = (t_ * cos) + (_rotate_half(t_) * sin)
+        return torch.cat((t_, t_pass_), dim=-1).type_as(t)
+
+
+class RMSNorm(torch.nn.Module):
+    def __init__(self, dim: int, eps: float = 1e-6):
+        super().__init__()
+        self.eps = eps
+        self.weight = nn.Parameter(torch.ones(dim))
+
+    def _norm(self, x):
+        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
+
+    def forward(self, x):
+        if rms_norm is not None and x.is_cuda:
+            return rms_norm(x, self.weight, self.eps)
+        else:
+            output = self._norm(x.float()).type_as(x)
+            return output * self.weight

quantize_config.json

@@ -0,0 +1,11 @@
+{
+  "bits": 4,
+  "group_size": 128,
+  "damp_percent": 0.01,
+  "desc_act": false,
+  "static_groups": false,
+  "sym": true,
+  "true_sequential": true,
+  "model_name_or_path": null,
+  "model_file_base_name": "model"
+}

qwen_generation_utils.py

@@ -0,0 +1,420 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Generation support."""
+
+from typing import Tuple, List, Union, Iterable
+
+import numpy as np
+import torch
+import torch.nn.functional as F
+from transformers import PreTrainedTokenizer
+from transformers import logging
+from transformers.generation import LogitsProcessor
+
+logger = logging.get_logger(__name__)
+
+# Types.
+HistoryType = List[Tuple[str, str]]
+TokensType = List[int]
+BatchTokensType = List[List[int]]
+
+
+def pad_batch(batch: BatchTokensType, pad_id: int, seq_length: int) -> BatchTokensType:
+    for tokens in batch:
+        context_length = len(tokens)
+        if context_length < seq_length:
+            tokens.extend([pad_id] * (seq_length - context_length))
+    return batch
+
+
+def get_ltor_masks_and_position_ids(
+    data,
+    eod_token,
+    reset_position_ids,
+    reset_attention_mask,
+    eod_mask_loss,
+):
+    """Build masks and position id for left to right model."""
+
+    # Extract batch size and sequence length.
+    micro_batch_size, seq_length = data.size()
+
+    # Attention mask (lower triangular).
+    if reset_attention_mask:
+        att_mask_batch = micro_batch_size
+    else:
+        att_mask_batch = 1
+    attention_mask = torch.tril(
+        torch.ones((att_mask_batch, seq_length, seq_length), device=data.device)
+    ).view(att_mask_batch, 1, seq_length, seq_length)
+
+    # Loss mask.
+    loss_mask = torch.ones(data.size(), dtype=torch.float, device=data.device)
+    if eod_mask_loss:
+        loss_mask[data == eod_token] = 0.0
+
+    # Position ids.
+    position_ids = torch.arange(seq_length, dtype=torch.long, device=data.device)
+    position_ids = position_ids.unsqueeze(0).expand_as(data)
+    # We need to clone as the ids will be modifed based on batch index.
+    if reset_position_ids:
+        position_ids = position_ids.clone()
+
+    if reset_position_ids or reset_attention_mask:
+        # Loop through the batches:
+        for b in range(micro_batch_size):
+
+            # Find indecies where EOD token is.
+            eod_index = position_ids[b, data[b] == eod_token]
+            # Detach indecies from positions if going to modify positions.
+            if reset_position_ids:
+                eod_index = eod_index.clone()
+
+            # Loop through EOD indecies:
+            prev_index = 0
+            for j in range(eod_index.size()[0]):
+                i = eod_index[j]
+                # Mask attention loss.
+                if reset_attention_mask:
+                    attention_mask[b, 0, (i + 1) :, : (i + 1)] = 0
+                # Reset positions.
+                if reset_position_ids:
+                    position_ids[b, (i + 1) :] -= i + 1 - prev_index
+                    prev_index = i + 1
+
+    # Convert attention mask to binary:
+    attention_mask = attention_mask < 0.5
+
+    return attention_mask, loss_mask, position_ids
+
+
+def get_batch(context_tokens: torch.LongTensor, eod_id: int):
+    """Generate batch from context tokens."""
+    # Move to GPU.
+    tokens = context_tokens.contiguous().to(context_tokens.device)
+    # Get the attention mask and postition ids.
+    attention_mask, _, position_ids = get_ltor_masks_and_position_ids(
+        tokens,
+        eod_id,
+        reset_position_ids=False,
+        reset_attention_mask=False,
+        eod_mask_loss=False,
+    )
+    return tokens, attention_mask, position_ids
+
+
+def get_stop_words_ids(chat_format, tokenizer):
+    if chat_format == "raw":
+        stop_words_ids = [tokenizer.encode("Human:"), [tokenizer.eod_id]]
+    elif chat_format == "chatml":
+        stop_words_ids = [[tokenizer.im_end_id], [tokenizer.im_start_id]]
+    else:
+        raise NotImplementedError(f"Unknown chat format {chat_format!r}")
+    return stop_words_ids
+
+
+def make_context(
+    tokenizer: PreTrainedTokenizer,
+    query: str,
+    history: List[Tuple[str, str]] = None,
+    system: str = "",
+    max_window_size: int = 6144,
+    chat_format: str = "chatml",
+):
+    if history is None:
+        history = []
+
+    if chat_format == "chatml":
+        im_start, im_end = "<|im_start|>", "<|im_end|>"
+        im_start_tokens = [tokenizer.im_start_id]
+        im_end_tokens = [tokenizer.im_end_id]
+        nl_tokens = tokenizer.encode("\n")
+
+        def _tokenize_str(role, content):
+            return f"{role}\n{content}", tokenizer.encode(
+                role, allowed_special=set(tokenizer.IMAGE_ST)
+            ) + nl_tokens + tokenizer.encode(content, allowed_special=set(tokenizer.IMAGE_ST))
+
+        system_text, system_tokens_part = _tokenize_str("system", system)
+        system_tokens = im_start_tokens + system_tokens_part + im_end_tokens
+
+        raw_text = ""
+        context_tokens = []
+
+        for turn_query, turn_response in reversed(history):
+            query_text, query_tokens_part = _tokenize_str("user", turn_query)
+            query_tokens = im_start_tokens + query_tokens_part + im_end_tokens
+            if turn_response is not None:
+                response_text, response_tokens_part = _tokenize_str(
+                    "assistant", turn_response
+                )
+                response_tokens = im_start_tokens + response_tokens_part + im_end_tokens
+
+                next_context_tokens = nl_tokens + query_tokens + nl_tokens + response_tokens
+                prev_chat = (
+                    f"\n{im_start}{query_text}{im_end}\n{im_start}{response_text}{im_end}"
+                )
+            else:
+                next_context_tokens = nl_tokens + query_tokens + nl_tokens
+                prev_chat = f"\n{im_start}{query_text}{im_end}\n"
+
+            current_context_size = (
+                len(system_tokens) + len(next_context_tokens) + len(context_tokens)
+            )
+            if current_context_size < max_window_size:
+                context_tokens = next_context_tokens + context_tokens
+                raw_text = prev_chat + raw_text
+            else:
+                break
+
+        context_tokens = system_tokens + context_tokens
+        raw_text = f"{im_start}{system_text}{im_end}" + raw_text
+        context_tokens += (
+            nl_tokens
+            + im_start_tokens
+            + _tokenize_str("user", query)[1]
+            + im_end_tokens
+            + nl_tokens
+            + im_start_tokens
+            + tokenizer.encode("assistant")
+            + nl_tokens
+        )
+        raw_text += f"\n{im_start}user\n{query}{im_end}\n{im_start}assistant\n"
+
+    elif chat_format == "raw":
+        raw_text = query
+        context_tokens = tokenizer.encode(raw_text)
+    else:
+        raise NotImplementedError(f"Unknown chat format {chat_format!r}")
+
+    return raw_text, context_tokens
+
+
+def _decode_default(
+    tokens: List[int],
+    *,
+    stop_words: List[str],
+    eod_words: List[str],
+    tokenizer: PreTrainedTokenizer,
+    raw_text_len: int,
+    verbose: bool = False,
+    return_end_reason: bool = False,
+    errors: str='replace',
+):
+    trim_decode_tokens = tokenizer.decode(tokens, errors=errors)[raw_text_len:]
+    if verbose:
+        print("\nRaw Generate: ", trim_decode_tokens)
+
+    end_reason = f"Gen length {len(tokens)}"
+    for stop_word in stop_words:
+        trim_decode_tokens = trim_decode_tokens.replace(stop_word, "").strip()
+    for eod_word in eod_words:
+        if eod_word in trim_decode_tokens:
+            end_reason = f"Gen {eod_word!r}"
+        trim_decode_tokens = trim_decode_tokens.split(eod_word)[0]
+    trim_decode_tokens = trim_decode_tokens.strip()
+    if verbose:
+        print("\nEnd Reason:", end_reason)
+        print("\nGenerate: ", trim_decode_tokens)
+
+    if return_end_reason:
+        return trim_decode_tokens, end_reason
+    else:
+        return trim_decode_tokens
+
+
+def _decode_chatml(
+    tokens: List[int],
+    *,
+    stop_words: List[str],
+    eod_token_ids: List[int],
+    tokenizer: PreTrainedTokenizer,
+    raw_text_len: int,
+    context_length: int,
+    verbose: bool = False,
+    return_end_reason: bool = False,
+    errors: str='replace'
+):
+    end_reason = f"Gen length {len(tokens)}"
+    eod_token_idx = context_length
+    for eod_token_idx in range(context_length, len(tokens)):
+        if tokens[eod_token_idx] in eod_token_ids:
+            end_reason = f"Gen {tokenizer.decode([tokens[eod_token_idx]])!r}"
+            break
+
+    trim_decode_tokens = tokenizer.decode(tokens[:eod_token_idx], errors=errors)[raw_text_len:]
+    if verbose:
+        print("\nRaw Generate w/o EOD:", tokenizer.decode(tokens, errors=errors)[raw_text_len:])
+        print("\nRaw Generate:", trim_decode_tokens)
+        print("\nEnd Reason:", end_reason)
+    for stop_word in stop_words:
+        trim_decode_tokens = trim_decode_tokens.replace(stop_word, "").strip()
+    trim_decode_tokens = trim_decode_tokens.strip()
+    if verbose:
+        print("\nGenerate:", trim_decode_tokens)
+
+    if return_end_reason:
+        return trim_decode_tokens, end_reason
+    else:
+        return trim_decode_tokens
+
+
+def decode_tokens(
+    tokens: Union[torch.LongTensor, TokensType],
+    tokenizer: PreTrainedTokenizer,
+    raw_text_len: int,
+    context_length: int,
+    chat_format: str,
+    verbose: bool = False,
+    return_end_reason: bool = False,
+    errors: str="replace",
+) -> str:
+    if torch.is_tensor(tokens):
+        tokens = tokens.cpu().numpy().tolist()
+
+    if chat_format == "chatml":
+        return _decode_chatml(
+            tokens,
+            stop_words=[],
+            eod_token_ids=[tokenizer.im_start_id, tokenizer.im_end_id],
+            tokenizer=tokenizer,
+            raw_text_len=raw_text_len,
+            context_length=context_length,
+            verbose=verbose,
+            return_end_reason=return_end_reason,
+            errors=errors,
+        )
+    elif chat_format == "raw":
+        return _decode_default(
+            tokens,
+            stop_words=["<|endoftext|>"],
+            eod_words=["<|endoftext|>"],
+            tokenizer=tokenizer,
+            raw_text_len=raw_text_len,
+            verbose=verbose,
+            return_end_reason=return_end_reason,
+            errors=errors,
+        )
+    else:
+        raise NotImplementedError(f"Unknown chat format {chat_format!r}")
+
+
+class StopWordsLogitsProcessor(LogitsProcessor):
+    """
+    :class:`transformers.LogitsProcessor` that enforces that when specified sequences appear, stop geration.
+
+    Args:
+        stop_words_ids (:obj:`List[List[int]]`):
+            List of list of token ids of stop ids. In order to get the tokens of the words
+            that should not appear in the generated text, use :obj:`tokenizer(bad_word,
+            add_prefix_space=True).input_ids`.
+        eos_token_id (:obj:`int`):
+            The id of the `end-of-sequence` token.
+    """
+
+    def __init__(self, stop_words_ids: Iterable[Iterable[int]], eos_token_id: int):
+
+        if not isinstance(stop_words_ids, List) or len(stop_words_ids) == 0:
+            raise ValueError(
+                f"`stop_words_ids` has to be a non-emtpy list, but is {stop_words_ids}."
+            )
+        if any(not isinstance(bad_word_ids, list) for bad_word_ids in stop_words_ids):
+            raise ValueError(
+                f"`stop_words_ids` has to be a list of lists, but is {stop_words_ids}."
+            )
+        if any(
+            any(
+                (not isinstance(token_id, (int, np.integer)) or token_id < 0)
+                for token_id in stop_word_ids
+            )
+            for stop_word_ids in stop_words_ids
+        ):
+            raise ValueError(
+                f"Each list in `stop_words_ids` has to be a list of positive integers, but is {stop_words_ids}."
+            )
+
+        self.stop_words_ids = list(
+            filter(
+                lambda bad_token_seq: bad_token_seq != [eos_token_id], stop_words_ids
+            )
+        )
+        self.eos_token_id = eos_token_id
+        for stop_token_seq in self.stop_words_ids:
+            assert (
+                len(stop_token_seq) > 0
+            ), "Stop words token sequences {} cannot have an empty list".format(
+                stop_words_ids
+            )
+
+    def __call__(
+        self, input_ids: torch.LongTensor, scores: torch.FloatTensor
+    ) -> torch.FloatTensor:
+        stopped_samples = self._calc_stopped_samples(input_ids)
+        for i, should_stop in enumerate(stopped_samples):
+            if should_stop:
+                scores[i, self.eos_token_id] = float(2**15)
+        return scores
+
+    def _tokens_match(self, prev_tokens: torch.LongTensor, tokens: List[int]) -> bool:
+        if len(tokens) == 0:
+            # if bad word tokens is just one token always ban it
+            return True
+        elif len(tokens) > len(prev_tokens):
+            # if bad word tokens are longer then prev input_ids they can't be equal
+            return False
+        elif prev_tokens[-len(tokens) :].tolist() == tokens:
+            # if tokens match
+            return True
+        else:
+            return False
+
+    def _calc_stopped_samples(self, prev_input_ids: Iterable[int]) -> Iterable[int]:
+        stopped_samples = []
+        for prev_input_ids_slice in prev_input_ids:
+            match = False
+            for stop_token_seq in self.stop_words_ids:
+                if self._tokens_match(prev_input_ids_slice, stop_token_seq):
+                    # if tokens do not match continue
+                    match = True
+                    break
+            stopped_samples.append(match)
+
+        return stopped_samples
+
+
+def top_k_logits(logits, top_k=0, top_p=0.0, filter_value=-float("Inf")):
+    """This function has been mostly taken from huggingface conversational
+    ai code at
+        https://medium.com/huggingface/how-to-build-a-state-of-the-art-
+             conversational-ai-with-transfer-learning-2d818ac26313"""
+
+    if top_k > 0:
+        # Remove all tokens with a probability less than the
+        # last token of the top-k
+        indices_to_remove = logits < torch.topk(logits, top_k)[0][..., -1, None]
+        logits[indices_to_remove] = filter_value
+
+    if top_p > 0.0:
+        # Cconvert to 1D
+        sorted_logits, sorted_indices = torch.sort(logits, descending=True, dim=-1)
+        cumulative_probs = torch.cumsum(F.softmax(sorted_logits, dim=-1), dim=-1)
+
+        # Remove tokens with cumulative probability above the threshold
+        sorted_indices_to_remove = cumulative_probs > top_p
+        # Shift the indices to the right to keep also the first token
+        # above the threshold
+        sorted_indices_to_remove[..., 1:] = sorted_indices_to_remove[..., :-1].clone()
+        sorted_indices_to_remove[..., 0] = 0
+        for i in range(sorted_indices.size(0)):
+            indices_to_remove = sorted_indices[i][sorted_indices_to_remove[i]]
+            logits[i][indices_to_remove] = filter_value
+
+    return logits
+
+
+def switch(val1, val2, boolean):
+    boolean = boolean.type_as(val1)
+    return (1 - boolean) * val1 + boolean * val2

requirements.txt

@@ -0,0 +1,10 @@
+transformers==4.32.0
+accelerate
+tiktoken
+einops
+transformers_stream_generator==0.0.4
+scipy
+torchvision
+pillow
+tensorboard
+matplotlib

tokenization_qwen.py

@@ -0,0 +1,584 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Tokenization classes for QWen."""
+
+import base64
+import logging
+import os
+import requests
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union, Any, Callable, Optional
+
+import tiktoken
+import numpy as np
+from PIL import Image
+from PIL import ImageFont
+from PIL import ImageDraw
+from transformers import PreTrainedTokenizer, AddedToken
+from transformers.utils import try_to_load_from_cache
+
+import matplotlib.colors as mcolors
+from matplotlib.font_manager import FontProperties
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken", "ttf": "SimSun.ttf"}
+
+PAT_STR = r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+"""
+ENDOFTEXT = "<|endoftext|>"
+IMSTART = "<|im_start|>"
+IMEND = "<|im_end|>"
+# as the default behavior is changed to allow special tokens in
+# regular texts, the surface forms of special tokens need to be
+# as different as possible to minimize the impact
+EXTRAS = tuple((f"<|extra_{i}|>" for i in range(205)))
+SPECIAL_TOKENS = (
+    ENDOFTEXT,
+    IMSTART,
+    IMEND,
+) + EXTRAS
+IMG_TOKEN_SPAN = 256
+
+
+def _load_tiktoken_bpe(tiktoken_bpe_file: str) -> Dict[bytes, int]:
+    with open(tiktoken_bpe_file, "rb") as f:
+        contents = f.read()
+    return {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in contents.splitlines() if line)
+    }
+
+def _list_find(
+    input_list: List[Any],
+    candidates: Tuple[Any],
+    start: int = 0,
+):
+    for i in range(start, len(input_list)):
+        if input_list[i] in candidates:
+            return i
+    return -1
+
+def _replace_closed_tag(
+    input_tokens: List[Any],
+    start_tags: Union[Any, Tuple[Any]],
+    end_tags: Union[Any, Tuple[Any]],
+    inclusive_replace_func: Callable,
+    exclusive_replace_func: Callable = lambda x: x,
+):
+    if isinstance(start_tags, (str, int)):
+        start_tags = (start_tags,)
+    if isinstance(end_tags, (str, int)):
+        end_tags = (end_tags,)
+    assert len(start_tags) == len(end_tags)
+
+    output_tokens = []
+    end = 0
+    while True:
+        start = _list_find(input_tokens, start_tags, end)
+        if start == -1:
+            break
+        output_tokens.extend(exclusive_replace_func(input_tokens[end : start]))
+        tag_idx = start_tags.index(input_tokens[start])
+        end = _list_find(input_tokens, (end_tags[tag_idx],), start)
+        if end == -1:
+            raise ValueError("Unclosed image token")
+        output_tokens.extend(inclusive_replace_func(input_tokens[start : end + 1]))
+        end += 1
+    output_tokens.extend(exclusive_replace_func(input_tokens[end : ]))
+    return output_tokens
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        image_start_tag='<img>',
+        image_end_tag='</img>',
+        image_pad_tag='<imgpad>',
+        ref_start_tag='<ref>',
+        ref_end_tag='</ref>',
+        box_start_tag='<box>',
+        box_end_tag='</box>',
+        quad_start_tag='<quad>',
+        quad_end_tag='</quad>',
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.image_start_tag = image_start_tag
+        self.image_end_tag = image_end_tag
+        self.image_pad_tag = image_pad_tag
+        self.ref_start_tag = ref_start_tag
+        self.ref_end_tag = ref_end_tag
+        self.box_start_tag = box_start_tag
+        self.box_end_tag = box_end_tag
+        self.quad_start_tag = quad_start_tag
+        self.quad_end_tag = quad_end_tag
+        self.IMAGE_ST = (
+            ref_start_tag, ref_end_tag,
+            box_start_tag, box_end_tag,
+            quad_start_tag, quad_end_tag,
+            image_start_tag, image_end_tag,
+            image_pad_tag
+        )
+
+        self.errors = errors  # how to handle errors in decoding
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in enumerate(
+                SPECIAL_TOKENS + self.IMAGE_ST, start=len(self.mergeable_ranks)
+            )
+        }
+        self.img_start_id = self.special_tokens[self.image_start_tag]
+        self.img_end_id = self.special_tokens[self.image_end_tag]
+        self.img_pad_id = self.special_tokens[self.image_pad_tag]
+        self.ref_start_id = self.special_tokens[self.ref_start_tag]
+        self.ref_end_id = self.special_tokens[self.ref_end_tag]
+        self.box_start_id = self.special_tokens[self.box_start_tag]
+        self.box_end_id = self.special_tokens[self.box_end_tag]
+        self.quad_start_id = self.special_tokens[self.quad_start_tag]
+        self.quad_end_id = self.special_tokens[self.quad_end_tag]
+
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        assert (
+            len(self.mergeable_ranks) + len(self.special_tokens) == enc.n_vocab
+        ), f"{len(self.mergeable_ranks) + len(self.special_tokens)} != {enc.n_vocab} in encoding"
+
+        self.decoder = {
+            v: k for k, v in self.mergeable_ranks.items()
+        }  # type: dict[int, bytes|str]
+        self.decoder.update({v: k for k, v in self.special_tokens.items()})
+
+        self.tokenizer = enc  # type: tiktoken.Encoding
+
+        self.eod_id = self.tokenizer.eot_token
+        self.im_start_id = self.special_tokens[IMSTART]
+        self.im_end_id = self.special_tokens[IMEND]
+
+    def __getstate__(self):
+        # for pickle lovers
+        state = self.__dict__.copy()
+        del state['tokenizer']
+        return state
+
+    def __setstate__(self, state):
+        # tokenizer is not python native; don't pass it; rebuild it
+        self.__dict__.update(state)
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        self.tokenizer = enc
+
+
+    def __len__(self) -> int:
+        return self.tokenizer.n_vocab
+
+    def get_vocab(self) -> Dict[bytes, int]:
+        return self.mergeable_ranks
+
+    def convert_tokens_to_ids(
+        self, tokens: Union[bytes, str, List[Union[bytes, str]]]
+    ) -> List[int]:
+        ids = []
+        if isinstance(tokens, (str, bytes)):
+            if tokens in self.special_tokens:
+                return self.special_tokens[tokens]
+            else:
+                return self.mergeable_ranks.get(tokens)
+        for token in tokens:
+            if token in self.special_tokens:
+                ids.append(self.special_tokens[token])
+            else:
+                ids.append(self.mergeable_ranks.get(token))
+        return ids
+
+    def _add_tokens(self, new_tokens: Union[List[str], List[AddedToken]], special_tokens: bool = False) -> int:
+        if not special_tokens and new_tokens:
+            raise ValueError('Adding regular tokens is not supported')
+        for token in new_tokens:
+            surface_form = token.content if isinstance(token, AddedToken) else token
+            if surface_form not in SPECIAL_TOKENS + self.IMAGE_ST:
+                raise ValueError('Adding unknown special tokens is not supported')
+        return 0
+
+    def save_vocabulary(self, save_directory: str, **kwargs) -> Tuple[str]:
+        """
+        Save only the vocabulary of the tokenizer (vocabulary).
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        file_path = os.path.join(save_directory, "qwen.tiktoken")
+        with open(file_path, "w", encoding="utf8") as w:
+            for k, v in self.mergeable_ranks.items():
+                line = base64.b64encode(k).decode("utf8") + " " + str(v) + "\n"
+                w.write(line)
+        return (file_path,)
+
+    def tokenize(
+        self,
+        text: str,
+        allowed_special: Union[Set, str] = "all",
+        disallowed_special: Union[Collection, str] = (),
+        **kwargs,
+    ) -> List[Union[bytes, str]]:
+        """
+        Converts a string in a sequence of tokens.
+
+        Args:
+            text (`str`):
+                The sequence to be encoded.
+            allowed_special (`Literal["all"]` or `set`):
+                The surface forms of the tokens to be encoded as special tokens in regular texts.
+                Default to "all".
+            disallowed_special (`Literal["all"]` or `Collection`):
+                The surface forms of the tokens that should not be in regular texts and trigger errors.
+                Default to an empty tuple.
+
+            kwargs (additional keyword arguments, *optional*):
+                Will be passed to the underlying model specific encode method.
+
+        Returns:
+            `List[bytes|str]`: The list of tokens.
+        """
+        tokens = []
+        text = unicodedata.normalize("NFC", text)
+
+        # this implementation takes a detour: text -> token id -> token surface forms
+        for t in self.tokenizer.encode(
+            text, allowed_special=allowed_special, disallowed_special=disallowed_special
+        ):
+            tokens.append(self.decoder[t])
+
+        def _encode_imgurl(img_tokens):
+            assert img_tokens[0] == self.image_start_tag and img_tokens[-1] == self.image_end_tag
+            img_tokens = img_tokens[1:-1]
+            img_url = b''.join(img_tokens)
+            out_img_tokens = list(map(self.decoder.get, img_url))
+            if len(out_img_tokens) > IMG_TOKEN_SPAN:
+                raise ValueError("The content in {}..{} is too long".format(
+                    self.image_start_tag, self.image_end_tag))
+            out_img_tokens.extend([self.image_pad_tag] * (IMG_TOKEN_SPAN - len(out_img_tokens)))
+            out_img_tokens = [self.image_start_tag] + out_img_tokens + [self.image_end_tag]
+            return out_img_tokens
+
+        return _replace_closed_tag(tokens, self.image_start_tag, self.image_end_tag, _encode_imgurl)
+
+    def convert_tokens_to_string(self, tokens: List[Union[bytes, str]]) -> str:
+        """
+        Converts a sequence of tokens in a single string.
+        """
+        text = ""
+        temp = b""
+        for t in tokens:
+            if isinstance(t, str):
+                if temp:
+                    text += temp.decode("utf-8", errors=self.errors)
+                    temp = b""
+                text += t
+            elif isinstance(t, bytes):
+                temp += t
+            else:
+                raise TypeError("token should only be of type types or str")
+        if temp:
+            text += temp.decode("utf-8", errors=self.errors)
+        return text
+
+    @property
+    def vocab_size(self):
+        return self.tokenizer.n_vocab
+
+    def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
+        """Converts an id to a token, special tokens included"""
+        if index in self.decoder:
+            return self.decoder[index]
+        raise ValueError("unknown ids")
+
+    def _convert_token_to_id(self, token: Union[bytes, str]) -> int:
+        """Converts a token to an id using the vocab, special tokens included"""
+        if token in self.special_tokens:
+            return self.special_tokens[token]
+        if token in self.mergeable_ranks:
+            return self.mergeable_ranks[token]
+        raise ValueError("unknown token")
+
+    def _tokenize(self, text: str, **kwargs):
+        """
+        Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based
+        vocabulary or sub-words for sub-word-based vocabularies (BPE/SentencePieces/WordPieces).
+
+        Do NOT take care of added tokens.
+        """
+        raise NotImplementedError
+
+    def _decode(
+        self,
+        token_ids: Union[int, List[int]],
+        skip_special_tokens: bool = False,
+        errors: str = None,
+        **kwargs,
+    ) -> str:
+        if isinstance(token_ids, int):
+            token_ids = [token_ids]
+
+        def _decode_imgurl(img_token_ids):
+            assert img_token_ids[0] == self.img_start_id and img_token_ids[-1] == self.img_end_id
+            img_token_ids = img_token_ids[1:-1]
+            img_token_ids = img_token_ids[ : img_token_ids.index(self.img_pad_id)]
+            img_url = bytes(img_token_ids).decode('utf-8')
+            return [self.img_start_id] + self.tokenizer.encode(img_url) + [self.img_end_id]
+
+        token_ids = _replace_closed_tag(token_ids, self.img_start_id, self.img_end_id, _decode_imgurl)
+
+        if skip_special_tokens:
+            token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)
+
+    def to_list_format(self, text: str):
+        text = unicodedata.normalize("NFC", text)
+        token_ids = self.tokenizer.encode(
+            text, allowed_special=set(self.IMAGE_ST + (ENDOFTEXT,)))
+
+        def _encode_vl_info(tokens):
+            if len(tokens) == 0:
+                return []
+            if tokens[0] == self.img_start_id and tokens[-1] == self.img_end_id:
+                key = 'image'
+            elif tokens[0] == self.ref_start_id and tokens[-1] == self.ref_end_id:
+                key = 'ref'
+            elif tokens[0] == self.box_start_id and tokens[-1] == self.box_end_id:
+                key = 'box'
+            elif tokens[0] == self.quad_start_id and tokens[-1] == self.quad_end_id:
+                key = 'quad'
+            else:
+                _tobytes = lambda x: x.encode('utf-8') if isinstance(x, str) else x
+                return [{'text': b''.join(map(_tobytes, map(self.decoder.get, tokens))).decode('utf-8')}]
+            _tobytes = lambda x: x.encode('utf-8') if isinstance(x, str) else x
+            val = b''.join(map(_tobytes, map(self.decoder.get, tokens[1:-1]))).decode('utf-8')
+            return [{key: val}]
+
+        return _replace_closed_tag(
+            token_ids,
+            (self.img_start_id, self.ref_start_id, self.box_start_id, self.quad_start_id),
+            (self.img_end_id, self.ref_end_id, self.box_end_id, self.quad_end_id),
+            _encode_vl_info,
+            _encode_vl_info,
+        )
+
+    def from_list_format(self, list_format: List[Dict]):
+        text = ''
+        num_images = 0
+        for ele in list_format:
+            if 'image' in ele:
+                num_images += 1
+                text += f'Picture {num_images}:'
+                text += self.image_start_tag + ele['image'] + self.image_end_tag
+                text += '\n'
+            elif 'text' in ele:
+                text += ele['text']
+            elif 'box' in ele:
+                if 'ref' in ele:
+                    text += self.ref_start_tag + ele['ref'] + self.ref_end_tag
+                for box in ele['box']:
+                    text += self.box_start_tag + '(%d,%d),(%d,%d)' % (box[0], box[1], box[2], box[3]) + self.box_end_tag
+            else:
+                raise ValueError("Unsupport element: " + str(ele))
+        return text
+
+    def _fetch_latest_picture(self, response, history):
+        if history is None:
+            history = []
+        _history = history + [(response, None)]
+        for q, r in _history[::-1]:
+            for ele in self.to_list_format(q)[::-1]:
+                if 'image' in ele:
+                    return ele['image']
+        return None
+
+    def _fetch_all_box_with_ref(self, text):
+        list_format = self.to_list_format(text)
+        output = []
+        for i, ele in enumerate(list_format):
+            if 'box' in ele:
+                bbox = tuple(map(int, ele['box'].replace('(', '').replace(')', '').split(',')))
+                assert len(bbox) == 4
+                output.append({'box': bbox})
+                if i > 0 and 'ref' in list_format[i-1]:
+                    output[-1]['ref'] = list_format[i-1]['ref'].strip()
+        return output
+
+    def draw_bbox_on_latest_picture(
+        self,
+        response,
+        history=None,
+    ) -> Optional[Image.Image]:
+        image = self._fetch_latest_picture(response, history)
+        if image is None:
+            return None
+        if image.startswith("http://") or image.startswith("https://"):
+            image = Image.open(requests.get(image, stream=True).raw).convert("RGB")
+            h, w = image.height, image.width
+        else:
+            image = np.asarray(Image.open(image).convert("RGB"))
+            h, w = image.shape[0], image.shape[1]
+        visualizer = Visualizer(image)
+
+        boxes = self._fetch_all_box_with_ref(response)
+        if not boxes:
+            return None
+        color = random.choice([_ for _ in mcolors.TABLEAU_COLORS.keys()]) # init color
+        for box in boxes:
+            if 'ref' in box: # random new color for new refexps
+                color = random.choice([_ for _ in mcolors.TABLEAU_COLORS.keys()])
+            x1, y1, x2, y2 = box['box']
+            x1, y1, x2, y2 = (int(x1 / 1000 * w), int(y1 / 1000 * h), int(x2 / 1000 * w), int(y2 / 1000 * h))
+            visualizer.draw_box((x1, y1, x2, y2), alpha=1, edge_color=color)
+            if 'ref' in box:
+                visualizer.draw_text(box['ref'], (x1, y1), color=color, horizontal_alignment="left")
+        return visualizer.output
+
+
+import colorsys
+import logging
+import math
+import numpy as np
+import matplotlib as mpl
+import matplotlib.colors as mplc
+import matplotlib.figure as mplfigure
+import torch
+from matplotlib.backends.backend_agg import FigureCanvasAgg
+from PIL import Image
+import random
+
+logger = logging.getLogger(__name__)
+
+
+class VisImage:
+    def __init__(self, img, scale=1.0):
+        self.img = img
+        self.scale = scale
+        self.width, self.height = img.shape[1], img.shape[0]
+        self._setup_figure(img)
+
+    def _setup_figure(self, img):
+        fig = mplfigure.Figure(frameon=False)
+        self.dpi = fig.get_dpi()
+        # add a small 1e-2 to avoid precision lost due to matplotlib's truncation
+        # (https://github.com/matplotlib/matplotlib/issues/15363)
+        fig.set_size_inches(
+            (self.width * self.scale + 1e-2) / self.dpi,
+            (self.height * self.scale + 1e-2) / self.dpi,
+        )
+        self.canvas = FigureCanvasAgg(fig)
+        # self.canvas = mpl.backends.backend_cairo.FigureCanvasCairo(fig)
+        ax = fig.add_axes([0.0, 0.0, 1.0, 1.0])
+        ax.axis("off")
+        self.fig = fig
+        self.ax = ax
+        self.reset_image(img)
+
+    def reset_image(self, img):
+        img = img.astype("uint8")
+        self.ax.imshow(img, extent=(0, self.width, self.height, 0), interpolation="nearest")
+
+    def save(self, filepath):
+        self.fig.savefig(filepath)
+
+    def get_image(self):
+        canvas = self.canvas
+        s, (width, height) = canvas.print_to_buffer()
+
+        buffer = np.frombuffer(s, dtype="uint8")
+
+        img_rgba = buffer.reshape(height, width, 4)
+        rgb, alpha = np.split(img_rgba, [3], axis=2)
+        return rgb.astype("uint8")
+
+
+class Visualizer:
+    def __init__(self, img_rgb, metadata=None, scale=1.0):
+        self.img = np.asarray(img_rgb).clip(0, 255).astype(np.uint8)
+        self.font_path = try_to_load_from_cache("Qwen/Qwen-VL-Chat", "SimSun.ttf")
+        self.output = VisImage(self.img, scale=scale)
+        self.cpu_device = torch.device("cpu")
+
+        # too small texts are useless, therefore clamp to 14
+        self._default_font_size = max(
+            np.sqrt(self.output.height * self.output.width) // 30, 15 // scale
+        )
+
+    def draw_text(
+        self,
+        text,
+        position,
+        *,
+        font_size=None,
+        color="g",
+        horizontal_alignment="center",
+        rotation=0,
+    ):
+        if not font_size:
+            font_size = self._default_font_size
+
+        # since the text background is dark, we don't want the text to be dark
+        color = np.maximum(list(mplc.to_rgb(color)), 0.2)
+        color[np.argmax(color)] = max(0.8, np.max(color))
+
+        x, y = position
+        self.output.ax.text(
+            x,
+            y,
+            text,
+            size=font_size * self.output.scale,
+            fontproperties=FontProperties(fname=self.font_path),
+            bbox={"facecolor": "black", "alpha": 0.8, "pad": 0.7, "edgecolor": "none"},
+            verticalalignment="top",
+            horizontalalignment=horizontal_alignment,
+            color=color,
+            zorder=10,
+            rotation=rotation,
+        )
+        return self.output
+
+    def draw_box(self, box_coord, alpha=0.5, edge_color="g", line_style="-"):
+        
+        x0, y0, x1, y1 = box_coord
+        width = x1 - x0
+        height = y1 - y0
+
+        linewidth = max(self._default_font_size / 4, 1)
+
+        self.output.ax.add_patch(
+            mpl.patches.Rectangle(
+                (x0, y0),
+                width,
+                height,
+                fill=False,
+                edgecolor=edge_color,
+                linewidth=linewidth * self.output.scale,
+                alpha=alpha,
+                linestyle=line_style,
+            )
+        )
+        return self.output
+
+    def get_output(self):
+        
+        return self.output

tokenizer_config.json

@@ -0,0 +1,10 @@
+{
+  "model_max_length": 8192,
+  "tokenizer_class": "QWenTokenizer",
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+      ]
+  }
+}

visual.py

@@ -0,0 +1,426 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from collections import OrderedDict
+import math
+import requests
+from io import BytesIO
+from functools import partial
+from PIL import Image
+from typing import Callable, Optional, Sequence, Tuple, List
+import numpy as np
+
+import torch
+from torch import nn
+from torch.nn import functional as F
+from torch.nn.init import trunc_normal_
+from torchvision import transforms
+from torchvision.transforms import InterpolationMode
+
+
+def get_abs_pos(abs_pos, tgt_size):
+    # abs_pos: L, C
+    # tgt_size: M
+    # return: M, C
+    src_size = int(math.sqrt(abs_pos.size(0)))
+    tgt_size = int(math.sqrt(tgt_size))
+    dtype = abs_pos.dtype
+
+    if src_size != tgt_size:
+        return F.interpolate(
+            abs_pos.float().reshape(1, src_size, src_size, -1).permute(0, 3, 1, 2),
+            size=(tgt_size, tgt_size),
+            mode="bicubic",
+            align_corners=False,
+        ).permute(0, 2, 3, 1).flatten(0, 2).to(dtype=dtype)
+    else:
+        return abs_pos
+
+# https://github.com/facebookresearch/mae/blob/efb2a8062c206524e35e47d04501ed4f544c0ae8/util/pos_embed.py#L20
+def get_2d_sincos_pos_embed(embed_dim, grid_size, cls_token=False):
+    """
+    grid_size: int of the grid height and width
+    return:
+    pos_embed: [grid_size*grid_size, embed_dim] or [1+grid_size*grid_size, embed_dim] (w/ or w/o cls_token)
+    """
+    grid_h = np.arange(grid_size, dtype=np.float32)
+    grid_w = np.arange(grid_size, dtype=np.float32)
+    grid = np.meshgrid(grid_w, grid_h)  # here w goes first
+    grid = np.stack(grid, axis=0)
+
+    grid = grid.reshape([2, 1, grid_size, grid_size])
+    pos_embed = get_2d_sincos_pos_embed_from_grid(embed_dim, grid)
+    if cls_token:
+        pos_embed = np.concatenate([np.zeros([1, embed_dim]), pos_embed], axis=0)
+    return pos_embed
+
+
+def get_2d_sincos_pos_embed_from_grid(embed_dim, grid):
+    assert embed_dim % 2 == 0
+
+    # use half of dimensions to encode grid_h
+    emb_h = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[0])  # (H*W, D/2)
+    emb_w = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[1])  # (H*W, D/2)
+
+    emb = np.concatenate([emb_h, emb_w], axis=1) # (H*W, D)
+    return emb
+
+
+def get_1d_sincos_pos_embed_from_grid(embed_dim, pos):
+    """
+    embed_dim: output dimension for each position
+    pos: a list of positions to be encoded: size (M,)
+    out: (M, D)
+    """
+    assert embed_dim % 2 == 0
+    omega = np.arange(embed_dim // 2, dtype=np.float32)
+    omega /= embed_dim / 2.
+    omega = 1. / 10000**omega  # (D/2,)
+
+    pos = pos.reshape(-1)  # (M,)
+    out = np.einsum('m,d->md', pos, omega)  # (M, D/2), outer product
+
+    emb_sin = np.sin(out) # (M, D/2)
+    emb_cos = np.cos(out) # (M, D/2)
+
+    emb = np.concatenate([emb_sin, emb_cos], axis=1)  # (M, D)
+    return emb
+
+
+class Resampler(nn.Module):
+    """
+    A 2D perceiver-resampler network with one cross attention layers by
+        (grid_size**2) learnable queries and 2d sincos pos_emb
+    Outputs:
+        A tensor with the shape of (grid_size**2, embed_dim)
+    """
+    def __init__(
+            self,
+            grid_size,
+            embed_dim,
+            num_heads,
+            kv_dim=None,
+            norm_layer=nn.LayerNorm
+    ):
+        super().__init__()
+        self.num_queries = grid_size ** 2
+        self.embed_dim = embed_dim
+        self.num_heads = num_heads
+
+        self.pos_embed = nn.Parameter(
+            torch.from_numpy(get_2d_sincos_pos_embed(embed_dim, grid_size)).float()
+        ).requires_grad_(False)
+
+        self.query = nn.Parameter(torch.zeros(self.num_queries, embed_dim))
+        trunc_normal_(self.query, std=.02)
+
+        if kv_dim is not None and kv_dim != embed_dim:
+            self.kv_proj = nn.Linear(kv_dim, embed_dim, bias=False)
+        else:
+            self.kv_proj = nn.Identity()
+
+        self.attn = nn.MultiheadAttention(embed_dim, num_heads)
+        self.ln_q = norm_layer(embed_dim)
+        self.ln_kv = norm_layer(embed_dim)
+        
+        self.apply(self._init_weights)
+
+    def _init_weights(self, m):
+        if isinstance(m, nn.Linear):
+            trunc_normal_(m.weight, std=.02)
+            if isinstance(m, nn.Linear) and m.bias is not None:
+                nn.init.constant_(m.bias, 0)
+        elif isinstance(m, nn.LayerNorm):
+            nn.init.constant_(m.bias, 0)
+            nn.init.constant_(m.weight, 1.0)
+
+    def forward(self, x, attn_mask=None):
+
+        pos_embed = get_abs_pos(self.pos_embed, x.size(1))
+
+        x = self.kv_proj(x)
+        x = self.ln_kv(x).permute(1, 0, 2)
+
+        N = x.shape[1]
+        q = self.ln_q(self.query)
+        out = self.attn(
+            self._repeat(q, N) + self.pos_embed.unsqueeze(1),
+            x + pos_embed.unsqueeze(1),
+            x,
+            attn_mask=attn_mask)[0]
+        return out.permute(1, 0, 2)
+
+    def _repeat(self, query, N: int):
+        return query.unsqueeze(1).repeat(1, N, 1)
+
+
+class VisualAttention(nn.Module):
+    """self-attention layer class.
+
+    Self-attention layer takes input with size [s, b, h]
+    and returns output of the same size.
+    """
+
+    def __init__(self, embed_dim, num_heads,
+                 bias=True, kdim=None, vdim=None):
+        super(VisualAttention, self).__init__()
+        self.embed_dim = embed_dim
+        self.kdim = kdim if kdim is not None else embed_dim
+        self.vdim = vdim if vdim is not None else embed_dim
+        self._qkv_same_embed_dim = self.kdim == embed_dim and self.vdim == embed_dim
+
+        self.num_heads = num_heads
+
+        # Per attention head and per partition values.
+        assert embed_dim % num_heads == 0
+        self.hidden_size_per_attention_head = embed_dim // num_heads
+        self.num_attention_heads_per_partition = num_heads
+        self.hidden_size_per_partition = embed_dim
+
+        # Strided linear layer.
+        assert self._qkv_same_embed_dim, 'Only Support SelfAttention Currently'
+        self.in_proj = nn.Linear(embed_dim, 3 * embed_dim)
+        self.out_proj = nn.Linear(embed_dim, embed_dim)
+        self.norm_factor = math.sqrt(self.hidden_size_per_attention_head)
+
+    def forward(self, query, key, value, attn_mask = None):
+        # query/key/value: [sq, b, h]
+        sq, b, _ = query.size()
+
+        assert query is key, 'Only Support Self-Attention Currently'
+        sk = sq
+        mixed_x_layer = self.in_proj(query)
+
+        # [sq, b, (np * 3 * hn)] --> [sq, b, np, 3 * hn]
+        new_tensor_shape = mixed_x_layer.size()[:-1] + \
+            (self.num_attention_heads_per_partition,
+             3 * self.hidden_size_per_attention_head)
+        mixed_x_layer = mixed_x_layer.view(*new_tensor_shape)
+
+        # [sq, b, np, 3 * hn] --> 3 [sq, b, np, hn]
+        query_layer, key_layer, value_layer = mixed_x_layer.split(
+            self.hidden_size_per_attention_head, dim=-1)
+
+        # [sq, b, np, hn] -> [sq, b * np, hn]
+        query_layer = query_layer.view(sq,
+            b * self.num_attention_heads_per_partition,
+            self.hidden_size_per_attention_head).transpose(0, 1)
+        # [sk, b, np, hn] -> [sk, b * np, hn]
+        key_layer = key_layer.view(sk,
+            b * self.num_attention_heads_per_partition,
+            self.hidden_size_per_attention_head).transpose(0, 1)
+
+        q_scaled = query_layer / self.norm_factor
+        if attn_mask is not None:
+            attention_probs = torch.baddbmm(attn_mask, q_scaled, key_layer.transpose(-2, -1))
+        else:
+            attention_probs = torch.bmm(q_scaled, key_layer.transpose(-2, -1))
+        attention_probs = attention_probs.softmax(dim=-1)
+
+        value_layer = value_layer.view(sk,
+            b * self.num_attention_heads_per_partition,
+            self.hidden_size_per_attention_head).transpose(0, 1)
+
+        # matmul: [b * np, sq, hn]
+        context_layer = torch.bmm(attention_probs, value_layer)
+
+        # change view [b, np, sq, hn]
+        context_layer = context_layer.view(b,
+            self.num_attention_heads_per_partition,
+            sq, self.hidden_size_per_attention_head)
+
+        # [b, np, sq, hn] --> [sq, b, np, hn]
+        context_layer = context_layer.permute(2, 0, 1, 3).contiguous()
+
+        # [sq, b, np, hn] --> [sq, b, hp]
+        new_context_layer_shape = context_layer.size()[:-2] + \
+            (self.hidden_size_per_partition,)
+        context_layer = context_layer.view(*new_context_layer_shape)
+
+        output = self.out_proj(context_layer)
+
+        return output
+
+
+class VisualAttentionBlock(nn.Module):
+    def __init__(
+            self,
+            d_model: int,
+            n_head: int,
+            mlp_ratio: float = 4.0,
+            act_layer: Callable = nn.GELU,
+            norm_layer: Callable = nn.LayerNorm,
+            is_cross_attention: bool = False,
+    ):
+        super().__init__()
+
+        self.ln_1 = norm_layer(d_model)
+        if is_cross_attention:
+            self.ln_1_kv = norm_layer(d_model)
+
+        self.ln_2 = norm_layer(d_model)
+        mlp_width = int(d_model * mlp_ratio)
+        self.attn = VisualAttention(d_model, n_head)
+        self.mlp = nn.Sequential(OrderedDict([
+            ("c_fc", nn.Linear(d_model, mlp_width)),
+            ("gelu", act_layer()),
+            ("c_proj", nn.Linear(mlp_width, d_model))
+        ]))
+
+    def attention(
+            self,
+            q_x: torch.Tensor,
+            k_x: Optional[torch.Tensor] = None,
+            v_x: Optional[torch.Tensor] = None,
+            attn_mask: Optional[torch.Tensor] = None,
+    ):
+        k_x = k_x if k_x is not None else q_x
+        v_x = v_x if v_x is not None else q_x
+
+        attn_mask = attn_mask.to(q_x.dtype) if attn_mask is not None else None
+        return self.attn(q_x, k_x, v_x, attn_mask=attn_mask)
+
+    def forward(
+            self,
+            q_x: torch.Tensor,
+            k_x: Optional[torch.Tensor] = None,
+            v_x: Optional[torch.Tensor] = None,
+            attn_mask: Optional[torch.Tensor] = None,
+    ):
+        k_x = self.ln_1_kv(k_x) if hasattr(self, "ln_1_kv") and k_x is not None else None
+        v_x = self.ln_1_kv(v_x) if hasattr(self, "ln_1_kv") and v_x is not None else None
+
+        x = q_x + self.attention(q_x=self.ln_1(q_x), k_x=k_x, v_x=v_x, attn_mask=attn_mask)
+        x = x + self.mlp(self.ln_2(x))
+        return x
+
+
+class TransformerBlock(nn.Module):
+    def __init__(
+            self,
+            width: int,
+            layers: int,
+            heads: int,
+            mlp_ratio: float = 4.0,
+            act_layer: Callable = nn.GELU,
+            norm_layer: Callable = nn.LayerNorm,
+    ):
+        super().__init__()
+        self.width = width
+        self.layers = layers
+
+        self.resblocks = nn.ModuleList([
+            VisualAttentionBlock(
+                width, heads, mlp_ratio, act_layer=act_layer, norm_layer=norm_layer)
+            for _ in range(layers)
+        ])
+
+    def get_cast_dtype(self) -> torch.dtype:
+        return self.resblocks[0].mlp.c_fc.weight.dtype
+
+    def get_cast_device(self) -> torch.device:
+        return self.resblocks[0].mlp.c_fc.weight.device
+
+    def forward(self, x: torch.Tensor, attn_mask: Optional[torch.Tensor] = None):
+        for r in self.resblocks:
+            x = r(x, attn_mask=attn_mask)
+        return x
+
+
+class VisionTransformer(nn.Module):
+
+    def __init__(
+            self,
+            image_size: int,
+            patch_size: int,
+            width: int,
+            layers: int,
+            heads: int,
+            mlp_ratio: float,
+            n_queries: int = 256,
+            output_dim: int = 512,
+            **kwargs
+    ):
+        super().__init__()
+        image_height, image_width = self.image_size = (image_size, image_size)
+        patch_height, patch_width = self.patch_size = (patch_size, patch_size)
+        self.grid_size = (image_height // patch_height, image_width // patch_width)
+        self.output_dim = output_dim
+
+        mean = (0.48145466, 0.4578275, 0.40821073)
+        std = (0.26862954, 0.26130258, 0.27577711)
+        self.image_transform = transforms.Compose([
+            transforms.Resize(
+                (image_size, image_size),
+                interpolation=InterpolationMode.BICUBIC
+            ),
+            transforms.ToTensor(),
+            transforms.Normalize(mean=mean, std=std),
+        ])
+
+        self.conv1 = nn.Conv2d(in_channels=3, out_channels=width, kernel_size=patch_size, stride=patch_size, bias=False)
+
+        # class embeddings and positional embeddings
+        scale = width ** -0.5
+        self.positional_embedding = nn.Parameter(scale * torch.randn(256, width))
+
+        norm_layer = partial(nn.LayerNorm, eps=1e-6)
+        act_layer = nn.GELU
+
+        self.ln_pre = norm_layer(width)
+        self.transformer = TransformerBlock(
+            width,
+            layers,
+            heads,
+            mlp_ratio,
+            act_layer=act_layer,
+            norm_layer=norm_layer,
+        )
+
+        self.attn_pool = Resampler(
+            grid_size=int(math.sqrt(n_queries)),
+            embed_dim=output_dim,
+            num_heads=output_dim // 128,
+            kv_dim=width,
+            norm_layer=norm_layer,
+        )
+        self.ln_post = norm_layer(output_dim)
+        self.proj = nn.Parameter((output_dim** -0.5) * torch.randn(output_dim, output_dim))
+
+    def forward(self, x: torch.Tensor):
+        x = x.to(
+            dtype=self.transformer.get_cast_dtype(),
+            device=self.transformer.get_cast_device(),
+        )
+        # to patches
+        x = self.conv1(x)  # shape = [*, width, grid, grid]
+        x = x.reshape(x.shape[0], x.shape[1], -1)  # shape = [*, width, grid ** 2]
+        x = x.permute(0, 2, 1)  # shape = [*, grid ** 2, width]
+
+        x = x + get_abs_pos(self.positional_embedding, x.size(1))
+
+        x = self.ln_pre(x)
+
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+
+        x = self.attn_pool(x)
+        x = self.ln_post(x)
+        x = x @ self.proj
+
+        return x
+
+    def encode(self, image_paths: List[str]):
+        images = []
+        for image_path in image_paths:
+            if image_path.startswith("http://") or image_path.startswith("https://"):
+                image = Image.open(requests.get(image_path, stream=True).raw)
+            else:
+                image = Image.open(image_path)
+            image = image.convert("RGB")
+            images.append(self.image_transform(image))
+        images = torch.stack(images, dim=0)
+        return self(images)