The CLI & python API for the well-known project gpt_academic.
- source installation.
bash init.bash
- Pip installation (Not recommended).
pip install void-terminal
- Chat
vt -a "hello, world!"
- Ask about how to do a linux command
vt -c "List all docker containers currently running on this system"
- Config (For all possible configurations, read
config.pyin the mother project.)
# Warning! This will write configuration into .bashrc and change your ENV variables !! Use with caution !! 警告,该命令会修改你的.bashrc文件,持久修改你的环境变量
vt --set_conf API_KEY "sk-123456789123456789123456789"
vt --set_conf LLM_MODEL "gpt-3.5-turbo"
vt --set_conf DEFAULT_WORKER_NUM "20"
- Chat without interaction
importvoid_terminalasvt# For more available configurations (including network proxy, api, using chatglm etc.),# see config.py of in the mother project:# https://github.com/binary-husky/gpt_academic.gitvt.set_conf(key="API_KEY", value="sk-xxxxxxxxxxxxxx") # or you can delete this line and set ENV variablevt.set_conf(key="LLM_MODEL", value="gpt-3.5-turbo") # or you can delete this line and set ENV variablechat_kwargs=vt.get_chat_default_kwargs()
chat_kwargs['inputs'] ='Hello, world!'result=vt.get_chat_handle()(**chat_kwargs)
print('\n*************\n'+result+'\n*************\n' )- Using mother project's plugin (Example: translate THIS readme file to Chinese)
importvoid_terminalasvtfromrich.liveimportLivefromrich.markdownimportMarkdown# vt.set_conf(key="API_URL_REDIRECT", value='{"https://api.openai.com/v1/chat/completions": "http://11.22.33.44:5566/v1/chat/completions"}')vt.set_conf(key="API_KEY", value="sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx")
vt.set_conf(key="LLM_MODEL", value="gpt-3.5-turbo")
plugin=vt.get_plugin_handle('void_terminal.crazy_functions.Markdown_Translate->TranslateMarkdownToSpecifiedLanguage')
plugin_kwargs=vt.get_plugin_default_kwargs()
plugin_kwargs['main_input'] ='./README.md'my_working_plugin=plugin(**plugin_kwargs)
withLive(Markdown(""), auto_refresh=False) aslive:
forcookies, chat, hist, msginmy_working_plugin:
md_str=vt.chat_to_markdown_str(chat)
md=Markdown(md_str)
live.update(md, refresh=True)- Using mother project's plugin (Example: chat with multiple LLM models)
importvoid_terminalasvtfromrich.liveimportLivefromrich.markdownimportMarkdownllm_model="gpt-3.5-turbo&gpt-4"vt.set_conf(key="API_KEY", value="sk-xxxxxxxxxxxxxx")
vt.set_conf(key="LLM_MODEL", value=llm_model)
plugin=vt.get_plugin_handle('void_terminal.crazy_functions.InquiryMultipleLargeLanguageModels->SimultaneousInquiry')
plugin_kwargs=vt.get_plugin_default_kwargs()
plugin_kwargs['main_input'] ='Hello, world!'plugin_kwargs['plugin_kwargs'] = {"advanced_arg": llm_model}
my_working_plugin=plugin(**plugin_kwargs)
withLive(Markdown(""), auto_refresh=False) aslive:
forcookies, chat, hist, msginmy_working_plugin:
md_str=vt.chat_to_markdown_str(chat)
md=Markdown(md_str)
live.update(md, refresh=True)SHOW_VOID_TERMINAL_LOGS. Set this Environment Variable to display void terminal logs