Releases: snakeying/GPT-Telegram-Bot
Releases · snakeying/GPT-Telegram-Bot
v1.0
🛠️ 系统优化与功能完善
- 问题修复:
- 修复了系统中发现的各种 bug,提高了整体稳定性
- /help 命令升级:
- 完善了 /help 命令的内容,提供更详细、更有用的帮助信息
- 本地化改进:
- 完成了大部分必要内容的中文化,提升了中文用户的使用体验
- 文档更新:
- 完善了 README 文件,为用户和开发者提供更全面的项目信息和指南
- 多语言支持规划:
- 下一阶段计划实现多语言支持,从英文版本开始
🛠️ System Optimization and Feature Enhancement
- Bug Fixes:
- Resolved various bugs discovered in the system, improving overall stability
- /help Command Upgrade:
- Enhanced the content of the /help command, providing more detailed and useful assistance
- Localization Improvement:
- Completed Chinese localization for most essential content, enhancing the experience for Chinese users
- Documentation Update:
- Refined the README file, offering more comprehensive project information and guidelines for users and developers
- Multilingual Support Planning:
- Next phase aims to implement multilingual support, starting with English version
v0.77
☁️ 新增 Azure OpenAI 模型支持
- Azure OpenAI API 集成:
- 新增 AZURE_OPENAI_API_KEY 环境变量:用于存储 Azure OpenAI 的 API 密钥
- 添加 AZURE_OPENAI_MODELS 环境变量:用于指定所需使用的 Azure OpenAI 模型
- Azure 端点配置:
- 新增 AZURE_OPENAI_ENDPOINT 环境变量:用于设置 Azure OpenAI API 的特定端点
- 输出功能支持:
- 实现 Azure OpenAI 的流式输出:提供实时、流畅的对话体验
- 支持 Markdown 格式输出:增强文本的结构化呈现和可读性
☁️ Azure OpenAI Model Support Addition
- Azure OpenAI API Integration:
- Added AZURE_OPENAI_API_KEY environment variable: For storing Azure OpenAI API key
- Introduced AZURE_OPENAI_MODELS environment variable: To specify desired Azure OpenAI models
- Azure Endpoint Configuration:
- Added AZURE_OPENAI_ENDPOINT environment variable: For setting the specific Azure OpenAI API endpoint
- Output Feature Support:
- Implemented Azure OpenAI's streaming output: Providing real-time, smooth conversation experience
- Markdown format output support: Enhancing structured text presentation and readability
v0.76
🧠 新增 Claude AI 支持
- Claude API 集成:
- 新增 CLAUDE_API_KEY 环境变量:用于存储 Claude 的 API 密钥
- 添加 CLAUDE_MODELS 环境变量:用于指定所需使用的 Claude 模型
- 自定义 API 端点:
- 新增 CLAUDE_ENDPOINT 环境变量:允许自定义 Claude API 的端点 URL
- 默认设置为 Anthropic 的官方端点
- 输出功能支持:
- 实现 Claude 的流式输出:提供实时、流畅的对话体验
- 支持 Markdown 格式输出:增强文本的结构化呈现和可读性
🧠 Claude AI Support Addition
- Claude API Integration:
- Added CLAUDE_API_KEY environment variable: For storing Claude's API key
- Introduced CLAUDE_MODELS environment variable: To specify desired Claude models
- Custom API Endpoint:
- Added CLAUDE_ENDPOINT environment variable: Allows customization of Claude API endpoint URL
- Default set to Anthropic's official endpoint
- Output Feature Support:
- Implemented Claude's streaming output: Providing real-time, smooth conversation experience
- Markdown format output support: Enhancing structured text presentation and readability
v0.75
🚀 新增 Groq 支持与系统优化
- Groq API 集成:
- 新增 GROQ_API_KEY 环境变量:用于认证 Groq API
- 添加 GROQ_MODELS 环境变量:支持多个 Groq 模型(以逗号分隔)
- Groq 输出特性:
- 不使用流式输出,确保响应的完整性
- 支持 Markdown 格式输出,提升文本的结构化呈现
- 系统优化:
- 将 Redis TTL 的过期时间延长至 1 天,提高会话管理效率
🚀 Groq Support Addition and System Optimization
- Groq API Integration:
- Added GROQ_API_KEY environment variable: For Groq API authentication
- Introduced GROQ_MODELS environment variable: Supports multiple Groq models (comma-separated)
- Groq Output Features:
- Non-streaming output to ensure response completeness
- Markdown format output support, enhancing structured text presentation
- System Enhancement:
- Extended Redis TTL expiration time to 1 day, improving session management efficiency
v0.72
🔄 上下文管理优化
- 移除固定上下文记录限制:
- 取消了之前硬编码的10条上下文记录限制
- 为用户提供更灵活的对话历史管理
- 引入 MAX_HISTORY_LENGTH 可选变量:
- 新增环境变量,允许自定义历史记录长度
- 默认值设置为50条记录
- 用户可根据个人需求自由调整,实现个性化的对话体验
🔄 Context Management Optimization
- Removal of Fixed Context Record Limit:
- Eliminated the previously hard-coded limit of 10 context records
- Provides users with more flexible dialogue history management
- Introduction of MAX_HISTORY_LENGTH Optional Variable:
- Added new environment variable to allow customization of history record length
- Default value set to 50 records
- Users can freely adjust based on personal needs, enabling a personalized conversation experience
v0.71
🔄 Gemini API 调整与优化
- 取消 Gemini 流式输出:
- 由于 Gemini API 对流式输出的支持不佳,我决定取消此功能
- 返回到标准输出模式,确保响应的稳定性和完整性
- Gemini Markdown 输出格式修正:
- 优化 Gemini 生成的 Markdown 内容的格式
- 确保文本结构、代码块、列表等元素的正确呈现
- 提高输出的可读性和一致性
🔄 Gemini API Adjustments and Optimizations
- Removal of Gemini Streaming Output:
- Due to poor support for streaming output in the Gemini API, I've decided to remove this feature
- Reverting to standard output mode to ensure response stability and completeness
- Correction of Gemini Markdown Output Format:
- Optimized the formatting of Markdown content generated by Gemini
- Ensured correct rendering of text structures, code blocks, lists, and other elements
- Improved readability and consistency of outputs
v0.7
🔧 功能优化:Gemini 集成与错误处理改进
- Gemini Markdown 输出支持:
- 为 Gemini API 响应添加 Markdown 格式支持
- 提升文本输出的结构化呈现和可读性
- 图片分析错误处理优化:
- 改进错误提示机制
- 在分析失败时返回 API 交互的详细结果(如果可用)
- 帮助用户更好地理解和解决潜在问题
🔧 Feature Enhancements: Gemini Integration and Error Handling Improvements
- Gemini Markdown Output Support:
- Added Markdown formatting support for Gemini API responses
- Enhances structured presentation and readability of text outputs
- Image Analysis Error Handling Optimization:
- Improved error notification mechanism
- Returns detailed API interaction results (if available) when analysis fails
- Helps users better understand and address potential issues
v0.65
🔄 新增功能:Gemini API 集成
- 新增可选环境变量:
- GEMINI_API_KEY:用于认证 Gemini API
- GOOGLE_MODELS:支持多个 Gemini 模型(以逗号分隔)
- GEMINI_ENDPOINT:允许自定义 API 端点
- Gemini 流式输出适配:实现实时、流畅的响应体验
- 命令适配:
- /new:支持 Gemini 模型的新对话启动
- /history:适配 Gemini 对话历史记录查看
- /switchmodel:支持在 Gemini 模型间切换
- 优化 /switchmodel 命令:
- 输入错误模型名称时,提供当前可用模型列表
🔄 New Feature: Gemini API Integration
- New optional environment variables:
- GEMINI_API_KEY: For Gemini API authentication
- GOOGLE_MODELS: Supports multiple Gemini models (comma-separated)
- GEMINI_ENDPOINT: Allows custom API endpoint specification
- Gemini streaming output adaptation: Enabling real-time, smooth response experience
- Command adaptations:
- /new: Supports initiating new conversations with Gemini models
- /history: Adapted for viewing Gemini conversation history
- /switchmodel: Supports switching between Gemini models
- Optimized /switchmodel command:
- Provides a list of currently available models when an incorrect model name is input
v0.6
📸 新功能:智能图片分析
- 新增 uploadhandler.js 文件,专门处理图片上传逻辑
- 支持的图片格式:'jpg', 'jpeg', 'png'。上传其他格式将返回错误提示
- 引入双重验证机制:使用 file-type 进行 MIME 类型检测,确保文件扩展名与实际类型匹配
- 仅支持多模态模型:gpt-4o、gpt-4o-mini 和 gpt-4
- 上传文件大小限制为 10MB
工作流程:用户上传文件 → Telegram Bot 接收 → 文件类型验证 → 扩展名一致性检查 → 相应处理
📸 New Feature: Intelligent Image Analysis
- Added uploadhandler.js file for dedicated image upload processing
- Supported image formats: 'jpg', 'jpeg', 'png'. Error message returned for other formats
- Implemented dual verification: Using file-type for MIME type detection, ensuring file extension matches actual type
- Supports multimodal models only: gpt-4o, gpt-4o-mini, and gpt-4
- File upload size limited to 10MB
Workflow: User uploads file → Telegram Bot receives → File type verification → Extension consistency check → Appropriate processing
v0.52
🚀 性能提升与个性化设置
- 引入流式响应功能:提供更快速、实时的对话体验
- 新增 SYSTEM_INIT_MESSAGE 环境变量:允许自定义 AI 助手的初始指令。默认为 "you are a helpful assistant."
- 添加 SYSTEM_INIT_MESSAGE_ROLE 变量:用于自定义初始化消息的角色,默认为 "system"
🚀 Performance Enhancement and Personalization
- Introduced streaming response: Providing faster, real-time conversation experience
- Added SYSTEM_INIT_MESSAGE environment variable: Allows customization of AI assistant's initial instructions. Defaults to "you are a helpful assistant."
- Implemented SYSTEM_INIT_MESSAGE_ROLE variable: For customizing the role of initialization message, default is "system"