什么是Textual Aesthetics，对于LLM生成的文本，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析
对于这个prompt："How can we reach neutron star by our hands?"生成的文本，
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.


对于这个prompt："How can we reach neutron star by our hands?"生成的文本，在保持文本语义质量不变的前提下，从文本美学的角度思考有哪些地方可以优化？
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.


"How can we optimize the text generated for the prompt 'How can we reach a neutron star by our hands?' from an aesthetic perspective while maintaining its semantic quality?"

It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant celestial objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.

Even if we could traverse such vast distances, approaching a neutron star would be perilous due to its intense gravitational pull and radiation. The conditions near a neutron star are far beyond what any human or known material could endure. Instead, we explore neutron stars using advanced telescopes and other scientific instruments.



对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析
对于这个prompt："How can we reach neutron star by our hands?"生成的文本，
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.

对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析
对于这个prompt："How can we reach neutron star by our hands?"生成的文本，
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.


对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析
对于这个prompt："How can we reach neutron star by our hands?"生成的文本，
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.


对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析，如果有可以修改的地方给出修改后的英语文本
对于这个prompt："Describe a possible solution to the environmental issue of air pollution."生成的文本，
One of the most effective solutions to the environmental issue of air pollution is promoting and investing in renewable energy sources. Traditional energy sources like coal and oil produce large amounts of greenhouse gases that contribute greatly to air pollution. Renewable energy sources like wind, solar, and hydropower sources do not produce greenhouse gases. They use clean energy sources that do not harm the environment and do not contribute to air pollution. \n\nAnother solution is improving our public transportation systems. Encouraging individuals to use public transportation, cycling, walking, or carpooling instead of their personal vehicles can greatly reduce emissions. This is especially helpful in urban areas where traffic congestion and subsequent air pollution is a common issue. \n\nReducing the use of single-use plastics can also be a significant step in reducing air pollution. The production, transportation, and disposal of plastic products contributes a large amount to greenhouse gases and air pollution. Using reusable bags, bottles, and containers can greatly reduce the amount of plastic waste and in turn, pollution.\n\nImplementing stricter regulations and penalties on industries that heavily pollute can also be a solution. Cleaner production practices can be enforced, and those who fail to comply can be penalized, raising awareness of the consequences of air pollution.\n\nFinally, increasing research and development in clean technologies can lead to innovative solutions to air pollution. This includes technologies that remove pollutants from the air, more efficient energy storage systems for renewable energy, and more effective waste management systems.\n\nThese solutions, when implemented in a systematic manner, can go a long way in reducing air pollution and protecting our environment.

对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析，如果有可以修改的地方给出修改后的英语文本
对于这个prompt："How can we reach neutron star by our hands?"生成的文本，
It is impossible to physically reach a neutron star with our hands. Neutron stars are incredibly distant astronomical objects, typically located light-years away from Earth. They are the remnants of massive stars that have undergone a supernova explosion and collapsed under their own gravity.\n\nEven if we could reach such a distance, approaching a neutron star would be extremely dangerous due to its intense gravity and radiation. The conditions near a neutron star are far beyond what any human or known material could withstand. Instead, we study neutron stars using advanced telescopes and other scientific instruments.


对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析，如果有可以修改的地方给出修改后的英语文本
对于这个prompt："Are the seeds of ostriches very big?"生成的文本，
Hello! I'm here to help answer your questions to the best of my ability. The seeds of ostriches are not particularly large compared to other birds. In fact, ostriches do not produce seeds at all, as they are a type of flightless bird and do not have the ability to produce flowers or fruits. Instead, ostriches lay eggs, which are quite large and are an important source of nutrition for the young birds. Is there anything else you would like to know about ostriches or eggs?

对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析，如果有可以修改的地方给出修改后的英语文本
对于这个prompt："I need to write some nodejs code that publishes a message to a Telegram group"生成的文本，
I can help you with that. First, you need to create a new bot on Telegram by following these steps:\n\n1. Open the Telegram app and search for the \"BotFather\" user.\n2. Start a chat with BotFather and send the command \"\/newbot\".\n3. Follow the instructions from BotFather to set a name and username for your bot.\n4. After creating the bot, BotFather will provide a token. Save this token as you'll need it later in the code.\n\nNow, you can use a package like `node-telegram-bot-api` to interact with the Telegram API. You can install this package using npm by running:\n\n```\nnpm install --save node-telegram-bot-api\n```\n\nHere's a basic example of how to send a message to a Telegram group using Node.js and the `node-telegram-bot-api` package. Remember to replace `<YOUR_TELEGRAM_BOT_TOKEN>` with the token provided by BotFather and `<GROUP_CHAT_ID>` with the chat ID of the group you want to send the message to.\n\n```javascript\nconst TelegramBot = require('node-telegram-bot-api');\n\n\/\/ Replace with your Telegram bot token\nconst token = '<YOUR_TELEGRAM_BOT_TOKEN>';\n\n\/\/ Create a new Telegram bot instance\nconst bot = new TelegramBot(token, { polling: true });\n\n\/\/ Replace with your target group chat ID (you can use '@your_group_name' if it's public)\nconst groupId = '<GROUP_CHAT_ID>';\n\n\/\/ Message you want to send\nconst message = 'Hello, Telegram group!';\n\n\/\/ Send the message using `sendMessage` method\nbot.sendMessage(groupId, message).then(() => {\n  console.log('Message sent to the group');\n}).catch((error) => {\n  console.error('Failed to send the message:', error);\n});\n```\n\nMake sure to replace the placeholders with the correct values. To get the group chat ID, you may need to add the bot to the group and have it listen for updates and log the chat ID when it receives a message.

对于LLM生成的文本，从文本的排版和布局以及文本结构对文本美学进行评估，可以从哪些方面评估文本的美观程度。对于下面的这个例子，给出具体的分析，如果有可以修改的地方给出修改后的英语文本
对于这个prompt："I need to write some nodejs code that publishes a message to a Telegram group"生成的文本，
I can help you with that. First, you need to create a new bot on Telegram by following these steps:\n\n1. Open the Telegram app and search for the \"BotFather\" user.\n2. Start a chat with BotFather and send the command \"\/newbot\".\n3. Follow the instructions from BotFather to set a name and username for your bot.\n4. After creating the bot, BotFather will provide a token. Save this token as you'll need it later in the code.\n\nNow, you can use a package like `node-telegram-bot-api` to interact with the Telegram API. You can install this package using npm by running:\n\n```\nnpm install --save node-telegram-bot-api\n```\n\nHere's a basic example of how to send a message to a Telegram group using Node.js and the `node-telegram-bot-api` package. Remember to replace `<YOUR_TELEGRAM_BOT_TOKEN>` with the token provided by BotFather and `<GROUP_CHAT_ID>` with the chat ID of the group you want to send the message to.\n\n```javascript\nconst TelegramBot = require('node-telegram-bot-api');\n\n\/\/ Replace with your Telegram bot token\nconst token = '<YOUR_TELEGRAM_BOT_TOKEN>';\n\n\/\/ Create a new Telegram bot instance\nconst bot = new TelegramBot(token, { polling: true });\n\n\/\/ Replace with your target group chat ID (you can use '@your_group_name' if it's public)\nconst groupId = '<GROUP_CHAT_ID>';\n\n\/\/ Message you want to send\nconst message = 'Hello, Telegram group!';\n\n\/\/ Send the message using `sendMessage` method\nbot.sendMessage(groupId, message).then(() => {\n  console.log('Message sent to the group');\n}).catch((error) => {\n  console.error('Failed to send the message:', error);\n});\n```\n\nMake sure to replace the placeholders with the correct values. To get the group chat ID, you may need to add the bot to the group and have it listen for updates and log the chat ID when it receives a message.

在LLM生成的文本中，我们经常需要对其进行美化处理。比如，在处理长文本时，我们需要合理分段、使用标题和子标题来组织内容，还需要通过列表、强调、加粗等方式来突出重点。此外，如果文本中包含代码或命令，应该使用代码块或行内代码来展示。  
  
为了确保文本的可读性，我们还应注意引用的使用，以及在适当的地方添加链接和表格。最后，保持整个文本风格的一致性也是非常重要的。

我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt，其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息，例如从标题分级、段落和换行、列表、强调和加粗、代码块和行内代码、引用、链接、表格、空白和间距、一致性等方面应该合理分布，使得到的文本更容易阅读，更容易理解其中的信息

system_template= \
"""You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective, determine whether the text needs modification, and then provide the revised text if necessary.  
  
*Format*   
  
Textual Aesthetic Analysis:
[Your analysis]

Does it need modification:
[Y/N]

Revised Text:
[Your revised text, if applicable]   
  
*Textual Aesthetic Elements to Consider*:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use headings and subheadings to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists and bullet points to break down complex information.  
5. **Markdown Formatting**: Apply markdown for bold, italic, and code blocks to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections and improve flow.  
"""
我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt，其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息，例如：
标题分级：合理使用Markdown的标题（#、##、###等）分级，确保层次清晰，内容结构明了。
段落和换行：适当地分段和换行，使文本不显得过于密集，保持阅读流畅。
列表和多级列表：对于需要列举的内容，使用有序列表（1. 2. 3.）或无序列表（- * +）进行整理
强调和加粗：合理使用斜体和加粗来强调重要信息。
代码块和行内代码：对于代码段，使用代码块，对于行内代码，使用行内代码。
引用：对于引用的内容，使用>符号。
链接：对于有链接的部分，使用Markdown的链接格式 描述。
表格：对于表格数据，使用Markdown表格格式。
空白和间距：适当地使用空行和空格，使文本不显得过于紧凑，保持适当的空白和间距。
一致性：确保整个文本的格式和风格一致。
脚注：如果需要使用脚注，使用[^1]格式，并在文本末尾添加脚注说明。
任务列表：对于任务或待办事项，使用任务列表格式 - [ ] 未完成 或 - [x] 已完成。
分隔线：在需要分隔内容的地方，使用---或***来插入分隔线。

我写了一个system promp 的template如下，按照上面的要求帮我对system prompt的template进行修改，给出修改后的英语的template，对于不需要修改的文本，给出""空字符串即可
system_template= \
"""You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective, determine whether the text needs modification, and then provide the revised text if necessary.  
  
*Format*   
  
Textual Aesthetic Analysis:
[Your analysis]

Does it need modification:
[Y/N]

Revised Text:
[Your revised text, if applicable]   
  
*Textual Aesthetic Elements to Consider*:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use headings and subheadings to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists and bullet points to break down complex information.  
5. **Markdown Formatting**: Apply markdown for bold, italic, and code blocks to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections and improve flow.  
"""
我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt，其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息。下面是我写的一个system prompt,帮我给出输出部分的修改格式建议，帮我对任务描述部分和要求部分进行适当的修改，强调对文本是否进行修改和是否添加每一个元素应当是合理的适当的而非必须的，目的是使生产的文本更加容易阅读理解，采用chain of thought的任务指导来帮助gpt更好的理解任务。给出修改后的英语template
system_template = """  
You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective, determine whether the text needs modification, and then provide the revised text if necessary.  
  
The goal is to ensure that the text, when rendered in Markdown or HTML, is visually appealing and easy to read and understand. These modifications should be reasonable and appropriate, rather than mandatory, and you may adjust or omit specific requirements as necessary.  
  
## Format  
  
**Textual Aesthetic Analysis:**  
Your analysis 
  
**Does it need modification:**  
If need modification: [[Y]] 
If donn't need modification: [[N]]  

  
**Revised Text:**  
[Your revised text, if applicable]  
  
## Textual Aesthetic Elements to Consider:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information.  
   - For unordered lists, use `-`, `*`, or `+`.  
   - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text.  
11. **Links**: Format hyperlinks using `[text](URL)`.  
12. **Tables**: Use Markdown tables for any tabular data.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""  

我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt，其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息，以下是我写的一个system prompt帮我给出输出部分的修改格式建议，以便能够方便的后续处理抽取出是否需要修改的Y或N和已经修改的文本，对于不需要修改的文本，给出""空字符串即可。给出修改后的英语template
system_template = """  
You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective, determine whether the text needs modification, and then provide the revised text if necessary.  
  
The goal is to ensure that the text, when rendered in Markdown or HTML, is visually appealing and easy to read and understand. These modifications should be reasonable and appropriate, rather than mandatory, and you may adjust or omit specific requirements as necessary.  
  
## Format  
  
**Textual Aesthetic Analysis:**  
Your analysis 
  
**Does it need modification:**  
If need modification: [[Y]] 
If donn't need modification: [[N]]  

  
**Revised Text:**  
[Your revised text, if applicable]  
  
## Textual Aesthetic Elements to Consider:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information.  
   - For unordered lists, use `-`, `*`, or `+`.  
   - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text.  
11. **Links**: Format hyperlinks using `[text](URL)`.  
12. **Tables**: Use Markdown tables for any tabular data.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""  

我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt，其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息，以下是我写的一个system prompt帮我给出输出部分的修改格式建议，以便能够方便的后续处理抽取出是否需要修改的Y或N和已经修改的文本，对于不需要修改的文本，给出""空字符串即可。给出修改后的英语template
user_template = """  
### User Instruction:  
{instruction}  
  
### Answer:  
{completion}  
  
Your task is to:  
  
1. Analyze the LLM-generated response from a textual aesthetic perspective.  
2. Determine whether the text needs modification.  
3. Provide a revised version of the text if necessary.  
  
*Format*   
  
**Textual Aesthetic Analysis:**  
[Your analysis]  
  
**Does it need modification:**  
If need modification: [[Y]]  
If don't need modification: [[N]]  
  
**Revised Text:**  
If need modification: [Your revised text]  
If don't need modification: [""]  
  
*Textual Aesthetic Elements to Consider*:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information.  
   - For unordered lists, use `-`, `*`, or `+`.  
   - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text.  
11. **Links**: Format hyperlinks using `[text](URL)`.  
12. **Tables**: Use Markdown tables for any tabular data.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
"""  

我需要写一个prompt，让gpt帮我将LLM对于给定instruction生成的文本在文本美学方面进行修改的prompt。其中文本美学主要指将LLM生成的文本用markdown的格式渲染出来后，在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息。下面是我写的一个system prompt,帮我给出输出部分的修改格式建议，帮我对任务描述部分和要求部分进行适当的修改，强调对文本是否进行修改和是否添加每一个元素应当是合理的适当的而非必须的，目的是使生产的文本更加容易阅读理解，采用chain of thought的任务指导来帮助gpt更好的理解任务。给出修改后的英语template
system_template = \
"""You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective, determine whether the text needs modification, and then provide the revised text if necessary.  
  
The goal is to ensure that the text, when rendered in Markdown, is visually appealing and easy to read and understand. These modifications should be reasonable and appropriate, rather than mandatory, and you may adjust or omit specific requirements as necessary.  
  
## Format  
  
**Textual Aesthetic Analysis:**  
Your analysis 
  
**Does it need modification:**  
If need modification: [[Y]] 
If donn't need modification: [[N]]  

  
**Revised Text:**  
[Your revised text, if applicable]  
  
## Textual Aesthetic Elements to Consider:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information.  
   - For unordered lists, use `-`, `*`, or `+`.  
   - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text.  
11. **Links**: Format hyperlinks using `[text](URL)`.  
12. **Tables**: Use Markdown tables for any tabular data.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""

You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Your job is to evaluate the provided user instruction and the LLM-generated response from an aesthetic perspective. Determine whether the text needs modification, and provide the revised text if necessary.   
  
The goal is to ensure that the text, when rendered in Markdown, is visually appealing, easy to read, and easy to understand. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.  
  
## Format  
  
**Textual Aesthetic Analysis:**    
Your analysis  
  
**Does it need modification:**    
If it needs modification: [[Y]]    
If it doesn't need modification: [[N]]  
  
**Revised Text:**    
[Your revised text, if applicable]  
  
## Textual Aesthetic Elements to Consider:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information.  
    - For unordered lists, use `-`, `*`, or `+`.  
    - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text.  
11. **Links**: Format hyperlinks using `[text](URL)`.  
12. **Tables**: Use Markdown tables for any tabular data.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  

system_template = \
"""You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Follow these steps to ensure a clear and structured approach:  
  
1. **Understand the Instruction**: First, read and understand the user instruction provided. This will help you grasp the context and purpose of the text.  
  
2. **Analyze the LLM-generated Response**: Carefully review the text generated by the LLM. Evaluate its structure, formatting, and overall readability.  
  
3. **Determine the Need for Modification**: Decide whether the text needs modification to improve its aesthetic quality. Ask yourself if the text is clear, well-organized, and easy to read. If the text is already satisfactory, no changes are necessary.  
  
4. **Identify Areas for Improvement**: If modification is needed, identify specific areas where improvements can be made. Consider elements like paragraph structure, headings, lists, and formatting.  
  
5. **Revise the Text**: Make appropriate modifications to enhance the text's visual appeal and readability. Apply changes judiciously and ensure they are reasonable and appropriate. Ensure that the revised text maintains a consistent style and format throughout.  
  
6. **Check Consistency**: Ensure that the revised text maintains a consistent style and format throughout.  
  
The goal is to ensure that the text, when rendered in Markdown, is visually appealing, easy to read, and easy to understand. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.  
  
## Format  
  
**Textual Aesthetic Analysis:**    
Your analysis  
  
**Does it need modification:**    
If it needs modification: [[Y]]    
If it doesn't need modification: [[N]]  
  
**Revised Text:**    
[Your revised text, if applicable]  
  
## Textual Aesthetic Elements to Consider:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text, if necessary.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability, if applicable.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information when needed.  
    - For unordered lists, use `-`, `*`, or `+`.  
    - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text, if necessary.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections when needed.  
7. **Consistency**: Maintain a consistent style throughout the document as appropriate.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow, if applicable.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points when necessary.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text if needed.  
11. **Links**: Format hyperlinks using `[text](URL)` when applicable.  
12. **Tables**: Use Markdown tables for any tabular data if required.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks if applicable.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document if necessary.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments when needed.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance, if necessary.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""

user_template = \
"""### User Instruction:   
{instruction}  
  
### Answer:   
{completion}  
  
Your task is to:  
  
1. **Analyze the LLM-generated response** from a textual aesthetic perspective. First, read and understand the text to grasp its context and purpose.  
2. **Determine whether the text needs modification**. Evaluate the text's structure, formatting, and readability. Decide if changes are necessary to improve its aesthetic quality.  
3. **Provide a revised version of the text if necessary**. Make reasonable and appropriate modifications to enhance the visual appeal and readability of the text.  
  
*Format*  
  
**Textual Aesthetic Analysis:**    
[Your analysis]  
  
**Does it need modification:**    
If it needs modification: [[Y]]    
If it doesn't need modification: [[N]]  
  
**Revised Text:**    
If it needs modification: [Your revised text]    
If it doesn't need modification: [""]  
  
*Textual Aesthetic Elements to Consider*:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text, if necessary.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability, if applicable.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information when needed.  
    - For unordered lists, use `-`, `*`, or `+`.  
    - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text, if necessary.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections when needed.  
7. **Consistency**: Maintain a consistent style throughout the document as appropriate.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow, if applicable.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points when necessary.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text if needed.  
11. **Links**: Format hyperlinks using `[text](URL)` when applicable.  
12. **Tables**: Use Markdown tables for any tabular data if required.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks if applicable.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document if necessary.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments when needed.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance, if necessary.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""


system_template = \
"""You are tasked with acting as a text rewriter to enhance the aesthetic quality of text generated by a Large Language Model (LLM). Follow these steps to ensure a clear and structured approach:  
  
1. Understand the Instruction: First, read and understand the user instruction provided. This will help you grasp the context and purpose of the text.  
  
2. Analyze the LLM-generated Response: Carefully review the text generated by the LLM. Evaluate its structure, formatting, and overall readability.  
  
3. Determine the Need for Modification: Decide whether the text needs modification to improve its aesthetic quality. Ask yourself if the text is clear, well-organized, and easy to read. If the text is already satisfactory, no changes are necessary.  
  
4. Identify Areas for Improvement: If modification is needed, identify specific areas where improvements can be made. Consider elements like paragraph structure, headings, lists, and formatting.  
  
5. Revise the Text: Make appropriate modifications to enhance the text's visual appeal and readability. Apply changes judiciously and ensure they are reasonable and appropriate.  
  
6. Check Consistency: Ensure that the revised text maintains a consistent style and format throughout.  
  
The goal is to ensure that the text, when rendered in Markdown, is visually appealing, easy to read, and easy to understand. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.  
"""

user_template = \
"""User Instruction:   
{instruction}  
  
Answer:   
{completion}  
  
Your task is to:  
  
1. Analyze the LLM-generated response from a textual aesthetic perspective. First, read and understand the text to grasp its context and purpose.  
2. Determine whether the text needs modification. Evaluate the text's structure, formatting, and readability. Decide if changes are necessary to improve its aesthetic quality.  
3. Provide a revised version of the text if necessary. Make reasonable and appropriate modifications to enhance the visual appeal and readability of the text.  
  
Format
  
Textual Aesthetic Analysis:    
[Your analysis]  
  
Does it need modification: 
If it needs modification: [[Y]]    
If it doesn't need modification: [[N]]  
  
Revised Text:
If it needs modification: [Your revised text]    
If it doesn't need modification: [""]  
  
Textual Aesthetic Elements to Consider:  
  
1. Paragraph Structure: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text, if necessary.  
2. Indentation: Apply consistent indentation where necessary.  
3. Headings and Subheadings: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability, if applicable.  
4. Lists and Bullet Points: Utilize lists (ordered and unordered) to break down complex information when needed.  
    - For unordered lists, use `-`, `*`, or `+`.  
    - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. Markdown Formatting: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text, if necessary.  
6. Line Spacing: Adjust line spacing to enhance readability. Use empty lines to separate sections when needed.  
7. Consistency: Maintain a consistent style throughout the document as appropriate.  
8. Visual Breaks: Use visual breaks (`---` or `***`) to separate different sections and improve flow, if applicable.  
9. Emphasis and Bold: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points when necessary.  
10. Blockquotes: Use blockquotes (`>`) for quotations or highlighted text if needed.  
11. Links: Format hyperlinks using `[text](URL)` when applicable.  
12. Tables: Use Markdown tables for any tabular data if required.  
13. Task Lists: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks if applicable.  
14. Footnotes: Use footnotes (`[^1]`) and provide the footnote content at the end of the document if necessary.  
15. Code and Inline Code: Use code blocks for larger code sections and inline code for shorter segments when needed.  
16. Whitespace and Spacing: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance, if necessary.  
  
These elements will ensure that the revised text is aesthetically pleasing, easy to read, and well-structured when rendered in Markdown.  
"""



user_template = \
""" 
<|User Prompt|>{instruction}  
   
<|The Start of Assistant's Answer|>    
{completion}    
<|The End of Assistant's Answer|>  
   
Your task is to:  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
   
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its aesthetic quality.  
    - Ask yourself if the text is clear, well-organized, and easy to read.  
    - If the text is already satisfactory, no changes are necessary.  
   
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's visual appeal and readability.  
    - Apply changes judiciously and ensure they are reasonable and appropriate.  
    - Ensure that the revised text maintains a consistent style and format throughout.  
   
**Textual Aesthetic Elements to Consider**:  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text if necessary.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use Markdown headings (`#`, `##`, `###`, etc.) to organize content and improve readability, if applicable.  
4. **Lists and Bullet Points**: Utilize lists (ordered and unordered) to break down complex information when needed.  
    - For unordered lists, use `-`, `*`, or `+`.  
    - For ordered lists, use `1.`, `2.`, `3.`, etc.  
5. **Markdown Formatting**: Apply Markdown for bold (`**bold**`), italic (`*italic*`), and code blocks (`` `code` `` or triple backticks for code blocks) to highlight important information or distinguish between different types of text, if necessary.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections when needed.  
7. **Consistency**: Maintain a consistent style throughout the document as appropriate.  
8. **Visual Breaks**: Use visual breaks (`---` or `***`) to separate different sections and improve flow, if applicable.  
9. **Emphasis and Bold**: Use bold (`**bold**`) and italic (`*italic*`) text to emphasize important points when necessary.  
10. **Blockquotes**: Use blockquotes (`>`) for quotations or highlighted text if needed.  
11. **Links**: Format hyperlinks using `[text](URL)` when applicable.  
12. **Tables**: Use Markdown tables for any tabular data if required.  
13. **Task Lists**: For task lists, use `- [ ]` for incomplete tasks and `- [x]` for completed tasks if applicable.  
14. **Footnotes**: Use footnotes (`[^1]`) and provide the footnote content at the end of the document if necessary.  
15. **Code and Inline Code**: Use code blocks for larger code sections and inline code for shorter segments when needed.  
16. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance, if necessary.  
   
**Format**    
**Textual Aesthetic Analysis**:  
- Your analysis  
   
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
   
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|> 
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
   
**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is generally clear, but it lacks emphasis on key points and could benefit from better structure. Adding bold text for emphasis and using bullet points to break down information will enhance readability.  
   
**Does it need modification**:  
[[Y]]  
   
**Revised Text**:  
<|Revised Content Start|>
**Summary of Project Benefits**  
   
The proposed project offers several key benefits:  
   
- **Increased Efficiency**: Implementing the new system will significantly reduce processing time.  
- **Cost Savings**: The project is expected to lower operational costs by 15%.  
- **Improved Accuracy**: Enhanced data validation processes will minimize errors.  
- **User Satisfaction**: User-friendly interfaces will improve overall satisfaction and usability.  
   
**Next Steps**  
   
To move forward, the following steps should be taken:  
   
1. **Approval**: Obtain necessary approvals from stakeholders.  
2. **Resources**: Allocate resources and assign team members.  
3. **Timeline**: Establish a detailed project timeline with milestones.  
4. **Implementation**: Begin the implementation phase with regular progress reviews.  
<|Revised Content End|>    
""" 

下面是我写的一个prompt，但使用这个prompt会让gpt对所有的文本增加一级标题，因为我们从LLM中得到的是回复而不是文章，目的是为了让文本更容易理解，以达到美化的效果，而不是所有的response都需要一级标题和段落列表来进行美化
请你思考如何写一个更好的prompt来帮助gpt对文本的结构进行必要的美观，然后根据你的思考检查我给出的prompt有哪些不合适的地方,对文本美学应该考虑的元素进行修改，然后给出修改意见
system_template = \
"""You are tasked with acting as a text rewriter to enhance the readability and comprehension of text generated by a Large Language Model (LLM). Your goal is to ensure the text is easy to read, easy to understand, and visually organized in a logical manner. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension."""

user_template = \
""" 
<|User Prompt|>{instruction}  
   
<|The Start of Assistant's Answer|>    
{completion}    
<|The End of Assistant's Answer|>  
   
Your task is to:  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
  
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its readability and comprehension.  
    - Ask yourself if the text is clear, well-organized, and easy to read.  
    - If the text is already satisfactory, no changes are necessary.  
  
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's readability and comprehension.  
    - Apply changes judiciously and ensure they are reasonable and appropriate.  
    - Ensure that the revised text maintains a consistent style and format throughout.   
   
**Textual Aesthetic Elements to Consider**:  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs to avoid dense text if necessary.  
2. **Indentation**: Apply consistent indentation where necessary.  
3. **Headings and Subheadings**: Use headings to organize content and improve readability, if applicable.  
4. **Lists and Bullet Points**: Utilize lists to break down complex information when needed.  
5. **Formatting for Emphasis**: Use bold or italic text to emphasize important points when necessary.  
6. **Line Spacing**: Adjust line spacing to enhance readability. Use empty lines to separate sections when needed.  
7. **Consistency**: Maintain a consistent style throughout the document as appropriate.  
8. **Visual Breaks**: Use visual breaks to separate different sections and improve flow, if applicable.  
9. **Blockquotes**: Use blockquotes for quotations or highlighted text if needed.  
10. **Links**: Format hyperlinks appropriately when applicable.  
11. **Tables**: Use tables for any tabular data if required.  
12. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance, if necessary.   
   
**Format**    
**Textual Aesthetic Analysis**:  
- Your analysis  
   
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
   
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|> 
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
   
**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.   
  
**Does it need modification**:  
[[N]]  
  
**Revised Text**:  
<|Revised Content Start|>""<|Revised Content End|>  
"""



system_template = \
"""You are tasked with acting as a text rewriter to enhance the readability and comprehension of text generated by a Large Language Model (LLM). Your goal is to ensure the text is easy to read, easy to understand, and visually organized in a logical manner. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.The text should also be helpful, relevant, and concise."""

user_template = """  
<|User Prompt|>{instruction}  
<|The Start of Assistant's Answer|>{completion}<|The End of Assistant's Answer|>  
  
Your task is to:  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its readability and comprehension.  
    - If the text is already satisfactory, no changes are necessary.  
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's readability and comprehension.  
    - Ensure the revised text maintains a consistent style and format throughout.  
    - Ensure the text is helpful, relevant, and concise:  
    
        - **Helpful**: Ensure the response accurately addresses the prompt or follows the instructions. If there's ambiguity or multiple interpretations, ask the user for clarification rather than making assumptions.  
        - **Relevant**: Ensure all parts of the response are directly connected to the prompt, avoiding unrelated details.  
        - **Concise**: Ensure the response is clear and to the point, without unnecessary verbosity. 
  
**Textual Aesthetic Elements to Consider**:  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs.  
2. **Indentation**: Apply consistent indentation if necessary.  
3. **Headings and Subheadings**: Use headings to organize content and improve readability, but only if the content naturally lends itself to such structure.  
4. **Lists and Bullet Points**: Utilize lists to break down complex information when applicable.  
5. **Formatting for Emphasis**: Use bold or italic text to emphasize important points judiciously.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections if applicable.  
9. **Blockquotes**: Use blockquotes for quotations or highlighted text.  
10. **Links**: Format hyperlinks appropriately when applicable.  
11. **Tables**: Use tables for any tabular data if required.  
12. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
**Format**:  
**Textual Aesthetic Analysis**:  
- Your analysis  
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|>  
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
  
**Example Output**:  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
**Does it need modification**: [[N]]  
**Revised Text**:  
<|Revised Content Start|>""<|Revised Content End|>  
"""

helpful, relevant, and concise. Helpful means the answer correctly responds to the prompt or follows the instructions. Note when user prompt has any ambiguity or more than one interpretation, it is more helpful and appropriate to ask for clarifications or more information from the user than providing an answer based on assumptions. Relevant means all parts of the response closely connect or are appropriate to what is being asked. Concise means the response is clear and not verbose or excessive.


以下是我用gpt对文本进行文本美学改写的prompt template，这prompt template的作用是将LLM对于给定instruction生成的文本进行文本美学方面的优化，使在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息。
system_template = \
"""You are tasked with acting as a text rewriter to enhance the readability and comprehension of text generated by a Large Language Model (LLM). Your goal is to ensure the text is easy to read, easy to understand, and visually organized in a logical manner. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.The text should also be helpful, relevant, and concise."""

user_template = """  
<|User Prompt|>{instruction}  
<|The Start of Assistant's Answer|>{completion}<|The End of Assistant's Answer|>  
  
Your task is to:  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its readability and comprehension.  
    - If the text is already satisfactory, no changes are necessary.  
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's readability and comprehension.  
    - Ensure the revised text maintains a consistent style and format throughout.  
    - Ensure the text is helpful, relevant, and concise:  
    
        - **Helpful**: Ensure the response accurately addresses the prompt or follows the instructions. If there's ambiguity or multiple interpretations, ask the user for clarification rather than making assumptions.  
        - **Relevant**: Ensure all parts of the response are directly connected to the prompt, avoiding unrelated details.  
        - **Concise**: Ensure the response is clear and to the point, without unnecessary verbosity. 
  
**Textual Aesthetic Elements to Consider**:  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs.  
2. **Indentation**: Apply consistent indentation if necessary.  
3. **Headings and Subheadings**: Use headings to organize content and improve readability, but only if the content naturally lends itself to such structure.  
4. **Lists and Bullet Points**: Utilize lists to break down complex information when applicable.  
5. **Formatting for Emphasis**: Use bold or italic text to emphasize important points judiciously.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections if applicable.  
9. **Blockquotes**: Use blockquotes for quotations or highlighted text.  
10. **Links**: Format hyperlinks appropriately when applicable.  
11. **Tables**: Use tables for any tabular data if required.  
12. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
**Format**:  
**Textual Aesthetic Analysis**:  
- Your analysis  
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|>  
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
  
**Example Output**:  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
**Does it need modification**: [[N]]  
**Revised Text**:  
<|Revised Content Start|>""<|Revised Content End|>  
"""
请你参考以下提供的2个对LLM生成的文本质量评分的prompt template，思考如何构造一个system prompt和user prompt，来对我上面改写的文本进行美学程度的打分，其中文本美学主要要求LLM生成的文本看起来更容易理解。
请给出你的思考构造过程，不要直接使用给出的两个评估的prompt中的要求，更多的从文本美观方面对文本进行打分,构造一个pairwise形式的评判judge，参考prompt template中使用cot的方式来对文本进行打分，最后给出对文本美学进行评分的英语system prompt template 和user prompt template
评分prompt1：
system_prompt = "A chat between a curious user and an artificial intelligence expert. The expert gives helpful, specific, and concise answers to the user's questions."

user_prompt = \
"""Given my answer to an instruction, your role is to provide specific and constructive feedback for me. You should find the best way for me to learn from your feedback and improve my performance. 

You should consider multiple aspects of my answer, including helpfulness, truthfulness, honesty, and to what extent the answer follows instructions.
---

### Instruction
{instruction}

### Answer
{completion}
---

Please act as a teacher and provide specific and constructive feedback. Besides describing the weaknesses of the answer, you should also provide specific suggestions to guide me toward understanding how to improve. Please note, however, that your suggestions should help me better complete the instructions, but you should not introduce new requirements that are not mentioned in the instructions. Your feedback should focus on enhancing my ability to think critically and respond accurately. However, never explicitly provide the reference answer, nor do polite phrases be required. Only respond with concise feedback in chat style. Finally, score the overall quality of the answer from 1 to 10, where 1 is the worst and 10 is the best.

*Format*
### Feedback
[Your feedback]
Overall Score: [1-10]

---

### Feedback
"""
评分prompt2：
system_prompt = "Please act as an impartial judge and evaluate the quality of the responses provided by two AI assistants to the user prompt displayed below. You will be given assistant A's answer and assistant B's answer. Your job is to evaluate which assistant's answer is better.\n\nBegin your evaluation by generating your own answer to the prompt. You must provide your answers before judging any answers.\n\nWhen evaluating the assistants' answers, compare both assistants' answers with your answer. You must identify and correct any mistakes or inaccurate information.\n\nThen consider if the assistant's answers are helpful, relevant, and concise. Helpful means the answer correctly responds to the prompt or follows the instructions. Note when user prompt has any ambiguity or more than one interpretation, it is more helpful and appropriate to ask for clarifications or more information from the user than providing an answer based on assumptions. Relevant means all parts of the response closely connect or are appropriate to what is being asked. Concise means the response is clear and not verbose or excessive.\n\nThen consider the creativity and novelty of the assistant's answers when needed. Finally, identify any missing important information in the assistants' answers that would be beneficial to include when responding to the user prompt.\n\nAfter providing your explanation, you must output only one of the following choices as your final verdict with a label:\n\n1. Assistant A is significantly better: [[A>>B]]\n2. Assistant A is slightly better: [[A>B]]\n3. Tie, relatively the same: [[A=B]]\n4. Assistant B is slightly better: [[B>A]]\n5. Assistant B is significantly better: [[B>>A]]\n\nExample output: \"My final verdict is tie: [[A=B]]\"."

user_template = '''["<|User Prompt|>\n{question_1}\n\n<|The Start of Assistant A's Answer|>\n{answer_1}\n<|The End of Assistant A's Answer|>\n\n<|The Start of Assistant B's Answer|>\n{answer_2}\n<|The End of Assistant B's Answer|>"]'''



A chat between a user and an AI assistant who is an expert in text aesthetics. The assistant's task is to evaluate the aesthetic quality of a given text and provide a score from 1 to 10. The evaluation should consider multiple aspects, including readability, visual organization, consistency, and overall structure. The assistant should provide constructive feedback and specific suggestions for improvement.  
  
---  
### Instruction  
{instruction}  
### Answer  
{completion}  
---  
Your role is to act as a text aesthetics expert and evaluate the given text based on the following criteria:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
### Feedback  
Provide specific and constructive feedback addressing the strengths and weaknesses of the text. Offer clear suggestions for improvement, focusing on enhancing readability, visual organization, consistency, and overall structure.  
  
Finally, score the overall aesthetic quality of the text from 1 to 10, where 1 is the worst and 10 is the best.  
  
*Format*  
### Feedback  
[Your feedback]  
Overall Score: [1-10]  

<|User Prompt|>  
{instruction}  
<|The Start of Assistant's Answer|>  
{completion}  
<|The End of Assistant's Answer|>  
  
Your task is to:  
  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
      
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its readability and comprehension.  
    - If the text is already satisfactory, no changes are necessary.  
  
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's readability and comprehension.  
    - Ensure the revised text maintains a consistent style and format throughout.  
    - Ensure the text is helpful, relevant, and concise:  
        - **Helpful**: Ensure the response accurately addresses the prompt or follows the instructions. If there's ambiguity or multiple interpretations, ask the user for clarification rather than making assumptions.  
        - **Relevant**: Ensure all parts of the response are directly connected to the prompt, avoiding unrelated details.  
        - **Concise**: Ensure the response is clear and to the point, without unnecessary verbosity.  
  
**Textual Aesthetic Elements to Consider**:  
  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs.  
2. **Indentation**: Apply consistent indentation if necessary.  
3. **Headings and Subheadings**: Use headings to organize content and improve readability, but only if the content naturally lends itself to such structure.  
4. **Lists and Bullet Points**: Utilize lists to break down complex information when applicable.  
5. **Formatting for Emphasis**: Use bold or italic text to emphasize important points judiciously.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections if applicable.  
9. **Blockquotes**: Use blockquotes for quotations or highlighted text.  
10. **Links**: Format hyperlinks appropriately when applicable.  
11. **Tables**: Use tables for any tabular data if required.  
12. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
**Format**:  
  
**Textual Aesthetic Analysis**:  
- Your analysis  
  
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
  
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|>  
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
  
**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
  
**Does it need modification**: [[N]]  
  
**Revised Text**:  
<|Revised Content Start|>""<|Revised Content End|>  


system_prompt = \
'''A chat between a user and an AI assistant who is an expert in text aesthetics. The assistant's task is to evaluate the aesthetic quality of a given text and provide a score from 1 to 10. The evaluation should consider multiple aspects, including readability, visual organization, consistency, and overall structure. The assistant should provide constructive feedback and specific suggestions for improvement.

---  
### Instruction  
{instruction}  
### Answer  
{completion}  
---  
Your role is to act as a text aesthetics expert and evaluate the given text based on the following criteria:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
### Feedback  
Provide specific and constructive feedback addressing the strengths and weaknesses of the text. Offer clear suggestions for improvement, focusing on enhancing readability, visual organization, consistency, and overall structure.  
  
Finally, score the overall aesthetic quality of the text from 1 to 10, where 1 is the worst and 10 is the best.  
  
*Format*  
### Feedback  
[Your feedback]  
Overall Score: [1-10]  
'''





       

user_prompt = \
'''
<|User Prompt|>  
{instruction}  
<|The Start of Assistant's Answer|>  
{completion}  
<|The End of Assistant's Answer|>  
  
Your task is to evaluate the aesthetic quality of the given text based on the following criteria:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
**Evaluation Steps**:  
  
1. **Analyze the Text**: Read and understand the text to grasp its context and purpose. Evaluate its structure, formatting, and overall readability.  
2. **Provide Feedback**: Offer specific and constructive feedback addressing the strengths and weaknesses of the text. Suggest clear improvements focusing on readability, visual organization, consistency, and overall structure.  
3. **Score the Text**: Finally, score the overall aesthetic quality of the text from 1 to 10, where 1 is the worst and 10 is the best.  
  
**Format**:  
  
**Textual Aesthetic Analysis**:  
- Your analysis  
  
**Feedback**:  
- Your feedback  
  
**Overall Score**: [1-10]  
  
**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
  
**Feedback**:  
The paragraphs are well-structured, but the text could benefit from more headings and bullet points to break down complex information.  
  
**Overall Score**: 8  
'''


<|User Prompt|>  
{instruction}  
<|The Start of Assistant's Answer|>  
{completion}  
<|The End of Assistant's Answer|>  
  
Your task is to evaluate the aesthetic quality of the given text based on the following criteria:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
**Evaluation Steps**:  
  
1. **Analyze the Text**: Read and understand the text to grasp its context and purpose. Evaluate its structure, formatting, and overall readability.  
2. **Provide Feedback**: Offer specific and constructive feedback addressing the strengths and weaknesses of the text. Suggest clear improvements focusing on readability, visual organization, consistency, and overall structure.  
3. **Score the Text**: Finally, score the overall aesthetic quality of the text from 1 to 10, where 1 is the worst and 10 is the best.  
  
**Format**:  
  
**Textual Aesthetic Analysis**:  
- Your analysis  
  
**Feedback**:  
- Your feedback  
  
**Overall Score**: [1-10]  
  
**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
  
**Feedback**:  
The paragraphs are well-structured, but the text could benefit from more headings and bullet points to break down complex information.  
  
**Overall Score**: 8  
'''

user_prompt = \
'''<|User Prompt|>  
{instruction}  
<|The Start of Assistant's Answer|>  
{completion}  
<|The End of Assistant's Answer|>  
  
Your task is to evaluate the aesthetic quality of the given text based on the following criteria:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
**Evaluation Steps**:  
  
1. **Analyze the Text**: Read and understand the text to grasp its context and purpose. Evaluate its structure, formatting, and overall readability.  
2. **Provide Feedback**: Offer specific and constructive feedback addressing the strengths and weaknesses of the text. Suggest clear improvements focusing on readability, visual organization, consistency, and overall structure.  
3. **Score the Text**: Finally, score the overall aesthetic quality of the text from 1 to 10, where 1 is the worst and 10 is the best.  
  
**Format**:  
  
**Textual Aesthetic Analysis**:  
- Your analysis  
  
**Feedback**:  
- Your feedback  
  
**Overall Score**: [1-10]  
'''

**Example Output**:  
  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
  
**Feedback**:  
The paragraphs are well-structured, but the text could benefit from more headings and bullet points to break down complex information.  
  
**Overall Score**: [8]  

以下是我用gpt对文本进行文本美学改写的prompt template，这prompt template的作用是将LLM对于给定instruction生成的文本进行文本美学方面的优化，使在观感上具有条理性，使得到的文本更容易阅读，更容易理解其中的信息。
system_template = \
"""You are tasked with acting as a text rewriter to enhance the readability and comprehension of text generated by a Large Language Model (LLM). Your goal is to ensure the text is easy to read, easy to understand, and visually organized in a logical manner. Modifications should be reasonable and appropriate, rather than mandatory. Each element should be used judiciously to enhance readability and comprehension.The text should also be helpful, relevant, and concise."""

user_template = """  
<|User Prompt|>{instruction}  
<|The Start of Assistant's Answer|>{completion}<|The End of Assistant's Answer|>  
  
Your task is to:  
1. **Analyze the LLM-generated response**:  
    - Read and understand the text to grasp its context and purpose.  
    - Carefully review the text generated by the LLM.  
    - Evaluate its structure, formatting, and overall readability.  
2. **Determine the Need for Modification**:  
    - Decide whether the text needs modification to improve its readability and comprehension.  
    - If the text is already satisfactory, no changes are necessary.  
3. **Provide a Revised Version of the Text if Necessary**:  
    - Make appropriate modifications to enhance the text's readability and comprehension.  
    - Ensure the revised text maintains a consistent style and format throughout.  
    - Ensure the text is helpful, relevant, and concise:  
    
        - **Helpful**: Ensure the response accurately addresses the prompt or follows the instructions. If there's ambiguity or multiple interpretations, ask the user for clarification rather than making assumptions.  
        - **Relevant**: Ensure all parts of the response are directly connected to the prompt, avoiding unrelated details.  
        - **Concise**: Ensure the response is clear and to the point, without unnecessary verbosity. 
  
**Textual Aesthetic Elements to Consider**:  
1. **Paragraph Structure**: Ensure paragraphs are of appropriate length and logically structured. Use appropriate spacing between paragraphs.  
2. **Indentation**: Apply consistent indentation if necessary.  
3. **Headings and Subheadings**: Use headings to organize content and improve readability, but only if the content naturally lends itself to such structure.  
4. **Lists and Bullet Points**: Utilize lists to break down complex information when applicable.  
5. **Formatting for Emphasis**: Use bold or italic text to emphasize important points judiciously.  
6. **Line Spacing**: Adjust line spacing to enhance readability.  
7. **Consistency**: Maintain a consistent style throughout the document.  
8. **Visual Breaks**: Use visual breaks to separate different sections if applicable.  
9. **Blockquotes**: Use blockquotes for quotations or highlighted text.  
10. **Links**: Format hyperlinks appropriately when applicable.  
11. **Tables**: Use tables for any tabular data if required.  
12. **Whitespace and Spacing**: Ensure appropriate use of whitespace and spacing to avoid a cluttered appearance.  
  
**Format**:  
**Textual Aesthetic Analysis**:  
- Your analysis  
**Does it need modification**:  
- If it needs modification: [[Y]]  
- If it doesn't need modification: [[N]]  
**Revised Text**:  
- If it needs modification: <|Revised Content Start|>Your revised text<|Revised Content End|>  
- If it doesn't need modification: <|Revised Content Start|>""<|Revised Content End|>  
  
**Example Output**:  
**Textual Aesthetic Analysis**:  
The text is clear, well-organized, and easy to read.  
**Does it need modification**: [[N]]  
**Revised Text**:  
<|Revised Content Start|>""<|Revised Content End|>  
"""
下面是我写的一个使用gpt基于给定instruction得到的两个不同LLM生成文本的judge prompt，思考如何构造一个system prompt和user prompt，来对我上面改写的文本进行美学程度的打分，其中文本美学主要要求LLM生成的文本看起来更容易理解。
请给出你的思考构造过程，更多的从文本美观方面对文本进行打分,构造一个pairwise形式的评判judge，prompt template中使用cot的方式来对文本进行打分，根据你自己的思考来对我给出的system prompt template 和user prompt template给出修改完事意见，最后给出修改后的英语template

## system prompt
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand.  
  
Begin your evaluation by considering the following aspects for each response:  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?   
  
After evaluating these aspects, compare both assistants' answers and make a final decision on which one is better in terms of textual aesthetics.  
  
You must output only one of the following choices as your final verdict with a label:  
1. Assistant A is significantly better: [[A>>B]]  
2. Assistant A is slightly better: [[A>B]]  
3. Tie, relatively the same: [[A=B]]  
4. Assistant B is slightly better: [[B>A]]  
5. Assistant B is significantly better: [[B>>A]]  
  
Example output: "My final verdict is Assistant A is slightly better: [[A>B]]."  

## user prompt
["<|User Prompt|>\n{question_1}\n\n<|The Start of Assistant A's Answer|>\n{answer_1}\n<|The End of Assistant A's Answer|>\n\n<|The Start of Assistant B's Answer|>\n{answer_2}\n<|The End of Assistant B's Answer|>"]  


## system prompt
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand.  
  
Begin your evaluation by considering the following aspects for each response:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
Follow these steps for your evaluation:  
1. **Analyze each response**: Carefully read and analyze both responses based on the criteria provided.  
2. **Compare both responses**: Determine which response excels in textual aesthetics considering all aspects.  
3. **Make a final decision**: Choose the response that is better in terms of textual aesthetics and justify your choice.  
  
You must output only one of the following choices as your final verdict with a label:  
1. Assistant A is significantly better: [[A>>B]]  
2. Assistant A is slightly better: [[A>B]]  
3. Tie, relatively the same: [[A=B]]  
4. Assistant B is slightly better: [[B>A]]  
5. Assistant B is significantly better: [[B>>A]]  
  
Example output: "My final verdict is Assistant A is slightly better: [[A>B]]."  

## user prompt  
"<|User Prompt|>\n{question_1}\n\n<|The Start of Assistant A's Answer|>\n{answer_1}\n<|The End of Assistant A's Answer|>\n\n<|The Start of Assistant B's Answer|>\n{answer_2}\n<|The End of Assistant B's Answer|>"  




下面是我写的一个使用gpt基于给定instruction得到的两个不同LLM生成文本的judge prompt，思考如何构造一个system prompt和user prompt，来对我上面改写的文本进行美学程度的打分，其中文本美学主要要求LLM生成的文本看起来更容易理解。
请给出你的思考构造过程，更多的从文本美观方面对文本进行打分,构造一个pairwise形式的评判judge，prompt template中使用cot的方式来对文本进行打分，根据你自己的思考来对我给出的system prompt template 和user prompt template给出修改完事意见，最后给出修改后的英语template

下面是我写的一个使用gpt基于给定instruction得到的两个不同LLM生成文本的judge prompt，来对文本进行美学程度的打分，其中文本美学主要要求LLM生成的文本看起来更容易理解。
现在我将两个LLM生成的文本将其渲染成图片，将图片传入其中，让LLM通过图片判断哪一个文本的文本美学更好，请你思考如何修改prompt，让gpt能够根据给出的图片进行打分，根据你的思考给出修改后的prompt
system_template = \
"""
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand.  
  
Begin your evaluation by considering the following aspects for each response:  
  
1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  
  
Follow these steps for your evaluation:  
1. **Analyze each response**: Carefully read and analyze both responses based on the criteria provided.  
2. **Compare both responses**: Determine which response excels in textual aesthetics considering all aspects.  
3. **Make a final decision**: Choose the response that is better in terms of textual aesthetics and justify your choice.  
  
You must output only one of the following choices as your final verdict with a label:  
1. Assistant A is significantly better: [[A>>B]]  
2. Assistant A is slightly better: [[A>B]]  
3. Tie, relatively the same: [[A=B]]  
4. Assistant B is slightly better: [[B>A]]  
5. Assistant B is significantly better: [[B>>A]]  
  
Example output: "My final verdict is Assistant A is slightly better: [[A>B]]."  
"""

## user prompt
user_template  = \
"""
<|User Prompt|>{instruction}
<|The Start of Assistant A's Answer|>
{answer_1}
<|The End of Assistant A's Answer|>

<|The Start of Assistant B's Answer|>
{answer_2}
<|The End of Assistant B's Answer|>"  
"""



system_template = \
"""
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. The responses are presented as images. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand, considering both textual and visual factors.

Begin your evaluation by considering the following aspects for each response:

1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?
3. **Consistency**: Does the text maintain a consistent style and format throughout?
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?
5. **Visual Clarity**: Is the text in the image clear and legible? Are the fonts and spacing visually appealing and easy on the eyes?

Follow these steps for your evaluation:
1. **Analyze each response**: Carefully examine both images based on the criteria provided.
2. **Compare both responses**: Determine which response excels in textual and visual aesthetics considering all aspects.
3. **Make a final decision**: Choose the response that is better in terms of textual and visual aesthetics and justify your choice.

You must output only one of the following choices as your final verdict with a label:
1. Assistant A is significantly better: [[A>>B]]
2. Assistant A is slightly better: [[A>B]]
3. Tie, relatively the same: [[A=B]]
4. Assistant B is slightly better: [[B>A]]
5. Assistant B is significantly better: [[B>>A]]

Example output: "My final verdict is Assistant A is slightly better: [[A>B]]."
"""

## user prompt
user_template  = \
"""
<|User Prompt|>{instruction}
<|The Start of Assistant A's Answer|>
![Image of Assistant A's Answer]({image_url_1})
<|The End of Assistant A's Answer|>

<|The Start of Assistant B's Answer|>
![Image of Assistant B's Answer]({image_url_2})
<|The End of Assistant B's Answer|>


对下面这个prompt进行修改，对user_template中给出instruction后，指出之后给出两种图片，第一张图片是Assistant A的答案，第二张图片是Assistant B的答案，给出修改后的英语prompt
system_template = \
"""
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. The responses are presented as images. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand, considering both textual and visual factors.

Begin your evaluation by considering the following aspects for each response:

1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?
3. **Consistency**: Does the text maintain a consistent style and format throughout?
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?
5. **Visual Clarity**: Is the text in the image clear and legible? Are the fonts and spacing visually appealing and easy on the eyes?

Follow these steps for your evaluation:
1. **Analyze each response**: Carefully examine both images based on the criteria provided.
2. **Compare both responses**: Determine which response excels in textual and visual aesthetics considering all aspects.
3. **Make a final decision**: Choose the response that is better in terms of textual and visual aesthetics and justify your choice.

You must output only one of the following choices as your final verdict with a label:
1. Assistant A is significantly better: [[A>>B]]
2. Assistant A is slightly better: [[A>B]]
3. Tie, relatively the same: [[A=B]]
4. Assistant B is slightly better: [[B>A]]
5. Assistant B is significantly better: [[B>>A]]

Example output: "My final verdict is Assistant A is slightly better: [[A>B]]."
"""

## user prompt
user_template  = \
"""
<|User Prompt|>{instruction}
<|The Start of Assistant A's Answer|>
![Image of Assistant A's Answer]({image_url_1})
<|The End of Assistant A's Answer|>

<|The Start of Assistant B's Answer|>
![Image of Assistant B's Answer]({image_url_2})
<|The End of Assistant B's Answer|>





下面是我写的一个使用gpt基于给定instruction得到的两个不同LLM生成文本的judge prompt，来对文本进行美学程度的打分，其中文本美学主要要求LLM生成的文本看起来更容易理解。我现在需要同时给出文本和图片对图片进行打分，帮我修改system prompt和user prompt，使得gpt能够根据给出的图片和文本进行打分，根据你的思考给出修改后的prompt
system_template_image = \
"""
You are an impartial judge tasked with evaluating the textual aesthetics of responses provided by two AI assistants to the user prompt displayed below. The responses are presented as images. Your goal is to determine which response is more aesthetically pleasing and easier to read and understand, considering both textual and visual factors.

Evaluate each response based on the following criteria:

1. **Readability**: Is the text easy to read and understand? Are the sentences of appropriate length and complexity?  
2. **Visual Organization**: Is the text visually organized in a logical manner? Are there appropriate headings, subheadings, lists, and other formatting elements?  
3. **Consistency**: Does the text maintain a consistent style and format throughout?  
4. **Overall Structure**: Are the paragraphs well-structured and logically connected? Is there appropriate spacing between paragraphs?  

Follow these steps for your evaluation:
1. **Analyze each response**: Carefully examine both images based on the criteria provided.
2. **Compare both responses**: Determine which response excels in textual and visual aesthetics considering all aspects.
3. **Make a final decision**: Choose the response that is better in terms of textual and visual aesthetics and justify your choice.

Output your final verdict with one of the following labels:
1. Assistant A is significantly better: [[A>>B]]
2. Assistant A is slightly better: [[A>B]]
3. Tie, relatively the same: [[A=B]]
4. Assistant B is slightly better: [[B>A]]
5. Assistant B is significantly better: [[B>>A]]

Example output:   
1. Analysis of Assistant A's response:  
    - Readability: ...  
    - Visual Organization: ...  
    - Consistency: ...  
    - Overall Structure: ...  
2. Analysis of Assistant B's response:  
    - Readability: ...  
    - Visual Organization: ...  
    - Consistency: ...  
    - Overall Structure: ...  
3. Comparison:  
    - Both responses are similar in readability, but...  
    - Assistant A has better visual organization...  
    - Assistant B's consistency is...  
    - Overall, Assistant A/B stands out in...
    
My final verdict is Assistant A is slightly better: [[A>B]].
"""

## user prompt
user_template_image  = \
"""
<|User Prompt|>{question_1}
Below are two images: the first one is Assistant A's response, and the second one is Assistant B's response. Please evaluate them based on the criteria provided and give the final verdict answer.
"""
## user prompt
user_template_image_v2  = \
"""
<|User Prompt|>{question_1}
<|The Start of Assistant A's Answer|>
{answer_1}
<|The End of Assistant A's Answer|>

<|The Start of Assistant B's Answer|>
{answer_2}
<|The End of Assistant B's Answer|>" 
Below are two images: the first one is Assistant A's response, and the second one is Assistant B's response. Please evaluate them based on the criteria provided and give the final verdict answer.
"""

messages = [
    {"role": "system", "content": f"{system}"},
    {"role": "user", "content": [
        {"type": "text", "text": f"{content}\n <|The Start of Assistant A's Answer|>\n"},
        {
            "image":base64_image1
        },
        {"type": "text", "text": "<|The End of Assistant A's Answer|>\n\n<|The Start of Assistant B's Answer|>\n"},
        {
            "image":base64_image2
        },
        {"type": "text", "text": "<|The End of Assistant B's Answer|>"},
    ]
        },
]






下面这个代码我在response中添加了一个middle值，帮我进行相同的处理

from typing import TYPE_CHECKING, Any, Dict, List, Optional, Sequence, Tuple

from ...extras.constants import IGNORE_INDEX
from ...extras.logging import get_logger
from .processor_utils import get_paligemma_token_type_ids, get_pixel_values, infer_seqlen


if TYPE_CHECKING:
    from transformers import PreTrainedTokenizer, ProcessorMixin

    from ...hparams import DataArguments
    from ..template import Template


logger = get_logger(__name__)


def _encode_pairwise_example(
    prompt: Sequence[Dict[str, str]],
    response: Sequence[Dict[str, str]],
    system: Optional[str],
    tools: Optional[str],
    template: "Template",
    tokenizer: "PreTrainedTokenizer",
    processor: Optional["ProcessorMixin"],
    cutoff_len: int,
) -> Tuple[List[int], List[int], List[int], List[int]]:
    if processor is not None and not hasattr(processor, "image_seq_length"):  # llava-like models
        prompt[0]["content"] = template.image_token + prompt[0]["content"]

    chosen_messages = prompt + [response[0]]
    rejected_messages = prompt + [response[1]]
    prompt_ids, chosen_ids = template.encode_oneturn(tokenizer, chosen_messages, system, tools)
    _, rejected_ids = template.encode_oneturn(tokenizer, rejected_messages, system, tools)

    if template.efficient_eos:
        chosen_ids += [tokenizer.eos_token_id]
        rejected_ids += [tokenizer.eos_token_id]

    if processor is not None and hasattr(processor, "image_seq_length"):  # paligemma models
        image_token_id = tokenizer.convert_tokens_to_ids(template.image_token)
        prompt_ids = [image_token_id] * getattr(processor, "image_seq_length") + prompt_ids

    # consider the response is more important
    source_len, target_len = infer_seqlen(len(prompt_ids), max(len(chosen_ids), len(rejected_ids)), cutoff_len)
    prompt_ids = prompt_ids[:source_len]
    chosen_ids = chosen_ids[:target_len]
    rejected_ids = rejected_ids[:target_len]

    chosen_input_ids = prompt_ids + chosen_ids
    chosen_labels = [IGNORE_INDEX] * source_len + chosen_ids
    rejected_input_ids = prompt_ids + rejected_ids
    rejected_labels = [IGNORE_INDEX] * source_len + rejected_ids

    return chosen_input_ids, chosen_labels, rejected_input_ids, rejected_labels


def preprocess_pairwise_dataset(
    examples: Dict[str, List[Any]],
    template: "Template",
    tokenizer: "PreTrainedTokenizer",
    processor: Optional["ProcessorMixin"],
    data_args: "DataArguments",
) -> Dict[str, List[List[int]]]:
    # build input pairs with format `<bos> X`, `Y1 <eos>` and `Y2 <eos>`
    model_inputs = {
        "chosen_input_ids": [],
        "chosen_attention_mask": [],
        "chosen_labels": [],
        "rejected_input_ids": [],
        "rejected_attention_mask": [],
        "rejected_labels": [],
    }
    if processor is not None:
        model_inputs["pixel_values"] = []
        if hasattr(processor, "image_seq_length"):  # paligemma models
            model_inputs["chosen_token_type_ids"] = []
            model_inputs["rejected_token_type_ids"] = []

    for i in range(len(examples["prompt"])):
        if len(examples["prompt"][i]) % 2 != 1 or len(examples["response"][i]) < 2:
            logger.warning("Dropped invalid example: {}".format(examples["prompt"][i] + examples["response"][i]))
            continue

        chosen_input_ids, chosen_labels, rejected_input_ids, rejected_labels = _encode_pairwise_example(
            prompt=examples["prompt"][i],
            response=examples["response"][i],
            system=examples["system"][i],
            tools=examples["tools"][i],
            template=template,
            tokenizer=tokenizer,
            processor=processor,
            cutoff_len=data_args.cutoff_len,
        )
        model_inputs["chosen_input_ids"].append(chosen_input_ids)
        model_inputs["chosen_attention_mask"].append([1] * len(chosen_input_ids))
        model_inputs["chosen_labels"].append(chosen_labels)
        model_inputs["rejected_input_ids"].append(rejected_input_ids)
        model_inputs["rejected_attention_mask"].append([1] * len(rejected_input_ids))
        model_inputs["rejected_labels"].append(rejected_labels)
        if processor is not None:
            model_inputs["pixel_values"].append(get_pixel_values(examples["images"][i], processor))
            if hasattr(processor, "image_seq_length"):  # paligemma models
                model_inputs["chosen_token_type_ids"].append(
                    get_paligemma_token_type_ids(len(chosen_input_ids), processor)
                )
                model_inputs["rejected_token_type_ids"].append(
                    get_paligemma_token_type_ids(len(rejected_input_ids), processor)
                )

    return model_inputs


def print_pairwise_dataset_example(example: Dict[str, List[int]], tokenizer: "PreTrainedTokenizer") -> None:
    valid_chosen_labels = list(filter(lambda x: x != IGNORE_INDEX, example["chosen_labels"]))
    valid_rejected_labels = list(filter(lambda x: x != IGNORE_INDEX, example["rejected_labels"]))
    print("chosen_input_ids:\n{}".format(example["chosen_input_ids"]))
    print("chosen_inputs:\n{}".format(tokenizer.decode(example["chosen_input_ids"], skip_special_tokens=False)))
    print("chosen_label_ids:\n{}".format(example["chosen_labels"]))
    print("chosen_labels:\n{}".format(tokenizer.decode(valid_chosen_labels, skip_special_tokens=False)))
    print("rejected_input_ids:\n{}".format(example["rejected_input_ids"]))
    print("rejected_inputs:\n{}".format(tokenizer.decode(example["rejected_input_ids"], skip_special_tokens=False)))
    print("rejected_label_ids:\n{}".format(example["rejected_labels"]))
    print("rejected_labels:\n{}".format(tokenizer.decode(valid_rejected_labels, skip_special_tokens=False)))
    
下面是response的值
response = [
    {"role": Role.ASSISTANT.value, "content": examples[dataset_attr.chosen][i]},
    {"role": Role.ASSISTANT.value, "content": examples[dataset_attr.midddle][i]},
    {"role": Role.ASSISTANT.value, "content": examples[dataset_attr.rejected][i]},
]