emu
1.0.0
使用Azure OpenAI API GPT-3.5-Turbo生成了用戶要求的程序描述的代碼。
代碼更改/添加以每月 - api request的基礎添加。
如果程序在運行時投擲，則將日誌錯誤發送到DEBUG的GPT，代碼或文本davinci模型。
Repo file structure:
.
├── config_dir #file configuration
├── config.json #project files, paths, token limits metadata
├── config.py #set by user API, model, temperature for each request
├── credentials #credentials folder for OpenAI API
│ └── self_config.py #Azure OpenAI API credentials & model names metadata
#Move sample_self_config.py to self_config.py and fill data
├── emu_cli.py #run this module to run the program
├── feature_common.py #common methods for feature requests to API
├── feature_manager.py #manager for each feature requested by user in the menu
├── ft_operations #non-API requests directory
│ ├── op_loadcode.py #loads code from local file to apply code change requests to it
│ ├── op_run_program.py #run the code
├── ft_requests #feature text requests directory
│ ├── feature_request_argparse.py #standard add argparse request
│ ├── feature_request_custom_req.py #user enters custom system and request prompt
│ ├── feature_request_debuglogs.py #send logs from running the program to API to debug error found in logs
│ ├── feature_request_docstrings.py #add docstrings
│ ├── feature_request_excpt_and_log.py #add exception handling and logs to the code
│ ├── feature_request_rawcode.py #generate initial code from a program description
├── log_list_handler.py
├── project #project output directory
│ ├── module.log
│ ├── module.py #code requested stored here and versioned
├── prompt_txt #prompt specs directory for each request
│ ├── clean_json_rq.py
│ ├── custom_req.py
│ ├── debug_rq.py
│ ├── docstrings_rq.py
│ ├── error_hndl_logging_rq.py
│ ├── input_and_argparse_rq.py
│ ├── raw_code_rq.py
├── README.md
├── requirements.txt
├── sample_self_config.py
├── tools #file and request management utilities directory
│ ├── file_management.py
│ └── request_utils.py
├── user_interaction.py #user interaction class
Python v3.10+必需的Python必需包裝：請參閱要求。
所需的路徑更改：將路徑更改到python環境的config.json：例如“ python_env_path”：“/home/home/sergio/sergio/anaconda3/bin/python“用於當前env設置值的路徑為“ python”的路徑
身份驗證：在此項目的根部創建“ CREDS”目錄，並保存在IT中sample_self_config.py。將此py文件重命名為self_config.py，然後輸入您的端點，模型/部署名稱和密鑰。使用Azure OpenAI API測試的項目。未經測試的Openai API。
根據請求配置OpenAI型號和溫度：
執行此程序：在命令行./emu_cli.py顯示菜單：
1. Generate Raw Code
Request model for code according to a description you provide.
2. Load Raw Code Script From File
3. Add Argparse
4. Exception Handling and Logging
5. User Custom Request
Requirement: code to be already loaded. Expected JSON response as specified in custom_req.py json_required_format variable.
6. Run Program And Request Repair of Debug Logs
Run the program and upon errors send the log error captured for the model to amend the code accordingly.
7. Add Docstrings To Program Code.
8. Set Menu Sequence
9. Run All
10. Exit
Choose your request:
切換以顯示每個請求的提示：在config_dir/config.py toggle bool show_request
如果您覺得有幫助，可以給我買咖啡:)
