gpt-academicc / docs /README.Korean.md
qingxu98's picture
update gr
47289f8
|
raw
history blame
25.3 kB

์ฐธ๊ณ 

์ด README๋Š” GPT ๋ฒˆ์—ญ์œผ๋กœ ์ƒ์„ฑ๋˜์—ˆ์Šต๋‹ˆ๋‹ค (์ด ํ”„๋กœ์ ํŠธ์˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์— ์˜ํ•ด ๊ตฌํ˜„๋จ) . 100% ์‹ ๋ขฐํ•  ์ˆ˜ ์—†์œผ๋ฏ€๋กœ ๋ฒˆ์—ญ ๊ฒฐ๊ณผ๋ฅผ ์ฃผ์˜ ๊นŠ๊ฒŒ ๊ฒ€ํ† ํ•˜์‹ญ์‹œ์˜ค.

2023.11.7: ์ข…์†์„ฑ์„ ์„ค์น˜ํ•  ๋•Œ, requirements.txt์— ์ง€์ •๋œ ๋ฒ„์ „์„ ์„ ํƒํ•˜์‹ญ์‹œ์˜ค. ์„ค์น˜ ๋ช…๋ น์–ด: pip install -r requirements.txt.

GPT ํ•™์ˆ  ์ตœ์ ํ™” (GPT Academic)

์ด ํ”„๋กœ์ ํŠธ๊ฐ€ ๋งˆ์Œ์— ๋“œ์‹ ๋‹ค๋ฉด, Star๋ฅผ ๋ถ€ํƒ๋“œ๋ฆฝ๋‹ˆ๋‹ค. ํŽธ๋ฆฌํ•œ ๋‹จ์ถ•ํ‚ค๋‚˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ๋ฐœ๊ฒฌํ•˜์…จ๋‹ค๋ฉด Pull Request๋ฅผ ํ™˜์˜ํ•ฉ๋‹ˆ๋‹ค! GPT๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ์ด ํ”„๋กœ์ ํŠธ๋ฅผ ์ž„์˜์˜ ์–ธ์–ด๋กœ ๋ฒˆ์—ญํ•˜๋ ค๋ฉด multi_language.py๋ฅผ ์ฝ๊ณ  ์‹คํ–‰ํ•˜์‹ญ์‹œ์˜ค (์‹คํ—˜์ ).

์ฐธ๊ณ 

  1. ๊ฐ•์กฐ ํ‘œ์‹œ๋œ ํ”Œ๋Ÿฌ๊ทธ์ธ (๋ฒ„ํŠผ)๋งŒ ํŒŒ์ผ์„ ์ฝ์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ผ๋ถ€ ํ”Œ๋Ÿฌ๊ทธ์ธ์€ ํ”Œ๋Ÿฌ๊ทธ์ธ ์˜์—ญ์˜ ๋“œ๋กญ๋‹ค์šด ๋ฉ”๋‰ด์— ์žˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ ์ƒˆ๋กœ์šด ํ”Œ๋Ÿฌ๊ทธ์ธ์— ๋Œ€ํ•œ ๋ชจ๋“  PR์„ ํ™˜์˜ํ•˜๋ฉฐ, ์ด๋ฅผ ๊ฐ€์žฅ ์šฐ์„ ์ ์œผ๋กœ ์ฒ˜๋ฆฌํ•ฉ๋‹ˆ๋‹ค.

  2. ์ด ํ”„๋กœ์ ํŠธ์˜ ๊ฐ ํŒŒ์ผ์˜ ๊ธฐ๋Šฅ์€ ์ž์ฒด ๋ถ„์„ ๋ณด๊ณ ์„œ self_analysis.md์—์„œ ์ž์„ธํžˆ ์„ค๋ช…๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค. ๋ฒ„์ „์ด ๋ฐ˜๋ณต๋จ์— ๋”ฐ๋ผ, ๊ด€๋ จ ๊ธฐ๋Šฅ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์–ธ์ œ๋“ ์ง€ ํด๋ฆญํ•˜์—ฌ GPT๋ฅผ ํ˜ธ์ถœํ•˜์—ฌ ํ”„๋กœ์ ํŠธ์˜ ์ž์ฒด ๋ถ„์„ ๋ณด๊ณ ์„œ๋ฅผ ๋‹ค์‹œ ์ƒ์„ฑํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ž์ฃผ ๋ฌป๋Š” ์งˆ๋ฌธ์€ ์œ„ํ‚ค๋ฅผ ์ฐธ์กฐํ•˜์‹ญ์‹œ์˜ค. ์ผ๋ฐ˜์ ์ธ ์„ค์น˜ ๋ฐฉ๋ฒ• | ์›ํด๋ฆญ ์„ค์น˜ ์Šคํฌ๋ฆฝํŠธ | ์„ค์ • ์„ค๋ช…์„œ

  1. ์ด ํ”„๋กœ์ ํŠธ๋Š” ChatGLM ๋“ฑ ๋Œ€ํ˜• ์–ธ์–ด ๋ชจ๋ธ (ChatGLM ๋“ฑ) ์‹คํ–‰์„ ์ง€์›ํ•˜๊ณ  ๊ถŒ์žฅํ•ฉ๋‹ˆ๋‹ค. ์—ฌ๋Ÿฌ ๊ฐœ์˜ API ํ‚ค๋ฅผ ๋™์‹œ์— ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ, ๊ตฌ์„ฑ ํŒŒ์ผ์— API_KEY="openai-key1,openai-key2,azure-key3,api2d-key4"์™€ ๊ฐ™์ด ์ž…๋ ฅํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. API_KEY๋ฅผ ์ผ์‹œ์ ์œผ๋กœ ๋ณ€๊ฒฝํ•ด์•ผ ํ•˜๋Š” ๊ฒฝ์šฐ, ์ž…๋ ฅ ์˜์—ญ์— ์ž„์‹œ API_KEY๋ฅผ ์ž…๋ ฅํ•œ ๋‹ค์Œ Enter ํ‚ค๋ฅผ ๋ˆ„๋ฅด๋ฉด ์ ์šฉ๋ฉ๋‹ˆ๋‹ค.
๊ธฐ๋Šฅ (โญ= ์ตœ๊ทผ ์ถ”๊ฐ€ ๊ธฐ๋Šฅ) ์„ค๋ช…
โญ์ƒˆ ๋ชจ๋ธ ์ถ”๊ฐ€! Baidu Qianfan์™€ Wenxin Yiyan, Tongyi Qianwen, Shanghai AI-Lab Shusheng, Xunfei Star, LLaMa2, Zhipu API, DALLE3
๋ฌธ์ฒด ๊ฐœ์„ , ๋ฒˆ์—ญ, ์ฝ”๋“œ ์„ค๋ช… ์ผ๊ด„์ ์ธ ๋ฌธ์ฒด ๊ฐœ์„ , ๋ฒˆ์—ญ, ๋…ผ๋ฌธ ๋ฌธ๋ฒ• ์˜ค๋ฅ˜ ํƒ์ƒ‰, ์ฝ”๋“œ ์„ค๋ช…
์‚ฌ์šฉ์ž ์ •์˜ ๋‹จ์ถ•ํ‚ค ์‚ฌ์šฉ์ž ์ •์˜ ๋‹จ์ถ•ํ‚ค ์ง€์›
๋ชจ๋“ˆํ™” ์„ค๊ณ„ ์‚ฌ์šฉ์ž ์ •์˜ ๊ฐ€๋Šฅํ•œ ๊ฐ•๋ ฅํ•œ ํ”Œ๋Ÿฌ๊ทธ์ธ ์ง€์›, ํ”Œ๋Ÿฌ๊ทธ์ธ ์ง€์› ํ•ซ ์—…๋ฐ์ดํŠธ
ํ”„๋กœ๊ทธ๋žจ ๋ถ„์„ [ํ”Œ๋Ÿฌ๊ทธ์ธ] ํ•œ ๋ฒˆ์— Python/C/C++/Java/Lua/... ํ”„๋กœ์ ํŠธ ํŠธ๋ฆฌ๋ฅผ ๋ถ„์„ํ•˜๊ฑฐ๋‚˜ ์ž์ฒด ๋ถ„์„
๋…ผ๋ฌธ ์ฝ๊ธฐ, ๋…ผ๋ฌธ ๋ฒˆ์—ญ [ํ”Œ๋Ÿฌ๊ทธ์ธ] LaTeX/PDF ๋…ผ๋ฌธ ์ „๋ฌธ์„ ์ฝ๊ณ  ์š”์•ฝ ์ƒ์„ฑ
LaTeX ์ „์ฒด ๋ฒˆ์—ญ, ๊ฐœ์„  [ํ”Œ๋Ÿฌ๊ทธ์ธ] LaTeX ๋…ผ๋ฌธ ๋ฒˆ์—ญ ๋˜๋Š” ๊ฐœ์„ 
์ผ๊ด„ ์ฃผ์„ ์ƒ์„ฑ [ํ”Œ๋Ÿฌ๊ทธ์ธ] ํ•จ์ˆ˜ ์ฃผ์„ ์ผ๊ด„ ์ƒ์„ฑ
Markdown ํ•œ / ์˜ ๋ฒˆ์—ญ ์œ„์˜ 5๊ฐœ ์–ธ์–ด๋กœ ์ž‘์„ฑ๋œ README๋ฅผ ์‚ดํŽด๋ณด์…จ๋‚˜์š”?
์ฑ„ํŒ… ๋ถ„์„ ๋ณด๊ณ ์„œ ์ƒ์„ฑ [ํ”Œ๋Ÿฌ๊ทธ์ธ] ์‹คํ–‰ ํ›„ ์š”์•ฝ ๋ณด๊ณ ์„œ ์ž๋™ ์ƒ์„ฑ
PDF ๋…ผ๋ฌธ ์ „์ฒด ๋ฒˆ์—ญ ๊ธฐ๋Šฅ [ํ”Œ๋Ÿฌ๊ทธ์ธ] PDF ๋…ผ๋ฌธ ์ œ๋ชฉ ๋ฐ ์š”์•ฝ ์ถ”์ถœ + ์ „์ฒด ๋ฒˆ์—ญ (๋ฉ€ํ‹ฐ ์Šค๋ ˆ๋“œ)
Arxiv ๋„์šฐ๋ฏธ [ํ”Œ๋Ÿฌ๊ทธ์ธ] arxiv ๋…ผ๋ฌธ url ์ž…๋ ฅ์‹œ ์š”์•ฝ ๋ฒˆ์—ญ + PDF ๋‹ค์šด๋กœ๋“œ
LaTeX ๋…ผ๋ฌธ ์ผ๊ด„ ๊ต์ • [ํ”Œ๋Ÿฌ๊ทธ์ธ] Grammarly๋ฅผ ๋ชจ์‚ฌํ•˜์—ฌ LaTeX ๋…ผ๋ฌธ์— ๋Œ€ํ•œ ๋ฌธ๋ฒ• ๋ฐ ๋งž์ถค๋ฒ• ์˜ค๋ฅ˜ ๊ต์ • + ๋Œ€์กฐ PDF ์ถœ๋ ฅ
Google ํ•™์ˆ  ํ†ตํ•ฉ ๋„์šฐ๋ฏธ ์ž„์˜์˜ Google ํ•™์ˆ  ๊ฒ€์ƒ‰ ํŽ˜์ด์ง€ URL์„ ์ง€์ •ํ•˜์—ฌ gpt๊ฐ€ related works๋ฅผ ์ž‘์„ฑํ•˜๊ฒŒ ํ•ด์ฃผ์„ธ์š”.
์ธํ„ฐ๋„ท ์ •๋ณด ์ง‘๊ณ„ + GPT [ํ”Œ๋Ÿฌ๊ทธ์ธ] ์ธํ„ฐ๋„ท์—์„œ ์ •๋ณด๋ฅผ ๊ฐ€์ ธ์™€์„œ ์งˆ๋ฌธ์— ๋Œ€๋‹ตํ•˜๋„๋ก GPT๋ฅผ ์ž๋™ํ™”ํ•˜์„ธ์š”. ์ •๋ณด๊ฐ€ ์ ˆ๋Œ€๋กœ ์˜ค๋ž˜๋˜์ง€ ์•Š๋„๋ก ํ•ด์ค๋‹ˆ๋‹ค.
โญArxiv ๋…ผ๋ฌธ ์„ธ์‹ฌํ•œ ๋ฒˆ์—ญ (Docker) [ํ”Œ๋Ÿฌ๊ทธ์ธ] arxiv ๋…ผ๋ฌธ์„ ๊ณ ํ’ˆ์งˆ ๋ฒˆ์—ญ์œผ๋กœ ๋ฒˆ์—ญํ•˜๋Š” ์ตœ๊ณ ์˜ ๋„๊ตฌ
โญ์‹ค์‹œ๊ฐ„ ์Œ์„ฑ ๋Œ€ํ™” ์ž…๋ ฅ [ํ”Œ๋Ÿฌ๊ทธ์ธ] ๋น„๋™๊ธฐ์ ์œผ๋กœ ์˜ค๋””์˜ค๋ฅผ ๋ชจ๋‹ˆํ„ฐ๋งํ•˜์—ฌ ๋ฌธ์žฅ์„ ์ž๋™์œผ๋กœ ๋ถ„์ ˆํ•˜๊ณ  ๋Œ€๋‹ต ์‹œ๊ธฐ๋ฅผ ์ž๋™์œผ๋กœ ์ฐพ์Šต๋‹ˆ๋‹ค.
์ˆ˜์‹/์ด๋ฏธ์ง€/ํ‘œ ํ‘œ์‹œ tex ํ˜•์‹ ๋ฐ ๋ Œ๋”๋ง ํ˜•์‹์˜ ์ˆ˜์‹์„ ๋™์‹œ์— ํ‘œ์‹œํ•˜๋ฉฐ, ์ˆ˜์‹ ๋ฐ ์ฝ”๋“œ ํ•˜์ด๋ผ์ดํŠธ ์ง€์›
โญAutoGen multi-agent ํ”Œ๋Ÿฌ๊ทธ์ธ [ํ”Œ๋Ÿฌ๊ทธ์ธ] Microsoft AutoGen์„ ํ™œ์šฉํ•˜์—ฌ ์—ฌ๋Ÿฌ ๊ฐœ์˜ ์—์ด์ „ํŠธ๊ฐ€ ์ง€๋Šฅ์ ์œผ๋กœ ๋ฐœ์ƒํ•˜๋Š” ๊ฐ€๋Šฅ์„ฑ์„ ํƒ์ƒ‰ํ•˜์„ธ์š”!
๋‹คํฌ ๋ชจ๋“œ ์ฃผ์ œ ์ง€์› ๋ธŒ๋ผ์šฐ์ €์˜ URL ๋’ค์— /?__theme=dark๋ฅผ ์ถ”๊ฐ€ํ•˜์—ฌ ๋‹คํฌ ๋ชจ๋“œ๋กœ ์ „ํ™˜ํ•˜์„ธ์š”.
๋‹ค์–‘ํ•œ LLM ๋ชจ๋ธ ์ง€์› GPT3.5, GPT4, Tsinghua ChatGLM2, Fudan MOSS์„ ํ•จ๊ป˜ ์‚ฌ์šฉํ•˜๋Š” ๋Š๋‚Œ์€ ์ข‹์„ ๊ฒƒ์ž…๋‹ˆ๋‹ค, ๊ทธ๋ ‡์ง€ ์•Š์Šต๋‹ˆ๊นŒ?
โญChatGLM2 fine-tuned ๋ชจ๋ธ ChatGLM2 fine-tuned ๋ชจ๋ธ ๋กœ๋“œ๋ฅผ ์ง€์›ํ•˜๋ฉฐ, ChatGLM2 fine-tuned ๋ณด์กฐ ํ”Œ๋Ÿฌ๊ทธ์ธ ์ œ๊ณต
๋” ๋งŽ์€ LLM ๋ชจ๋ธ ์—ฐ๊ฒฐ, huggingface ๋ฐฐํฌ ์ง€์› Newbing ์ธํ„ฐํŽ˜์ด์Šค(์‹  ๋ฐ), Tsinghua Jittorllms ๋„์ž…, LLaMA์™€ Pangu-alpha๋ฅผ ์ง€์›ํ•ฉ๋‹ˆ๋‹ค.
โญvoid-terminal ํŒจํ‚ค์ง€ GUI์—์„œ ๋…๋ฆฝ, Python์—์„œ ์ด ํ”„๋กœ์ ํŠธ์˜ ๋ชจ๋“  ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์ง์ ‘ ํ˜ธ์ถœ (๊ฐœ๋ฐœ ์ค‘)
โญVoid ํ„ฐ๋ฏธ๋„ ํ”Œ๋Ÿฌ๊ทธ์ธ [ํ”Œ๋Ÿฌ๊ทธ์ธ] ์ž์—ฐ์–ด๋กœ ์ด ํ”„๋กœ์ ํŠธ์˜ ๋‹ค๋ฅธ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์ง์ ‘ ์˜์†ํ•ฉ๋‹ˆ๋‹ค.
๊ธฐํƒ€ ์ƒˆ๋กœ์šด ๊ธฐ๋Šฅ ์†Œ๊ฐœ (์ด๋ฏธ์ง€ ์ƒ์„ฑ ๋“ฑ) โ€ฆโ€ฆ ๋ณธ ๋ฌธ์„œ ๋งจ ๋ ์ฐธ์กฐ โ€ฆโ€ฆ
  • ์ƒˆ๋กœ์šด ์ธํ„ฐํŽ˜์ด์Šค(config.py์˜ LAYOUT ์˜ต์…˜ ์ˆ˜์ •์œผ๋กœ "์™ผ์ชฝ-์˜ค๋ฅธ์ชฝ ๋ ˆ์ด์•„์›ƒ"๊ณผ "์œ„-์•„๋ž˜ ๋ ˆ์ด์•„์›ƒ"์„ ์ „ํ™˜ํ•  ์ˆ˜ ์žˆ์Œ)

  • ๋ชจ๋“  ๋ฒ„ํŠผ์€ functional.py๋ฅผ ๋™์ ์œผ๋กœ ์ฝ์–ด ์ƒ์„ฑ๋˜๋ฏ€๋กœ ์›ํ•˜๋Š”๋Œ€๋กœ ์‚ฌ์šฉ์ž ์ •์˜ ๊ธฐ๋Šฅ์„ ์ถ”๊ฐ€ํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ ํด๋ฆฝ ๋ณด๋“œ๋ฅผ ํ•ด์ œํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

  • ๋ฌธ์ฒด ๊ฐœ์„ /์˜ค๋ฅ˜ ์ˆ˜์ •

  • If the output contains equations, they will be displayed in both tex format and rendered format for easy copying and reading.

  • Don't feel like looking at the project code? Just give it to ChatGPT and let it dazzle you.

  • Mix and match multiple powerful language models (ChatGLM + OpenAI-GPT3.5 + API2D-GPT4)

Installation

Installation Method I: Run Directly (Windows, Linux or MacOS)

  1. Download the project
git clone --depth=1 https://github.com/binary-husky/gpt_academic.git
cd gpt_academic
  1. Configure API_KEY

In config.py, configure the API KEY and other settings, click here to view special network environment configuration methods. Wiki pageใ€‚

" The program will first check if there is a confidential configuration file named config_private.py and use its configuration to override the configuration with the same name in config.py. If you can understand this reading logic, we strongly recommend that you create a new configuration file named config_private.py next to config.py and move (copy) the configuration from config.py to config_private.py (only copy the modified configuration items). "

" You can configure the project through environment variables. The format of the environment variables can be found in the docker-compose.yml file or our Wiki page. The priority of the configuration reading is: environment variables > config_private.py > config.py. "

  1. Install dependencies
# (Option I: if familiar with python, python>=3.9) Note: Use the official pip source or Aliyun pip source. Temporary switching source method: python -m pip install -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/
python -m pip install -r requirements.txt

# (Option II: using Anaconda) The steps are similar (https://www.bilibili.com/video/BV1rc411W7Dr):
conda create -n gptac_venv python=3.11    # Create an Anaconda environment
conda activate gptac_venv                 # Activate the Anaconda environment
python -m pip install -r requirements.txt # This step is the same as the pip installation step
Click here to expand if you need support for Tsinghua ChatGLM2/Fudan MOSS/RWKV backend

[Optional Step] If you need support for Tsinghua ChatGLM2/Fudan MOSS as the backend, you need to install additional dependencies (Prerequisites: Familiar with Python + Have used Pytorch + Sufficient computer configuration):

# [Optional Step I] Support for Tsinghua ChatGLM2. Note for Tsinghua ChatGLM: If you encounter the error "Call ChatGLM fail cannot load ChatGLM parameters", refer to the following: 1: The default installation above is torch+cpu version. To use cuda, uninstall torch and reinstall torch+cuda; 2: If you cannot load the model due to insufficient computer configuration, you can modify the model precision in request_llm/bridge_chatglm.py, change AutoTokenizer.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True) to AutoTokenizer.from_pretrained("THUDM/chatglm-6b-int4", trust_remote_code=True)
python -m pip install -r request_llms/requirements_chatglm.txt

# [Optional Step II] Support for Fudan MOSS
python -m pip install -r request_llms/requirements_moss.txt
git clone --depth=1 https://github.com/OpenLMLab/MOSS.git request_llms/moss  # When executing this line of code, make sure you are in the project root path

# [Optional Step III] Support for RWKV Runner
Refer to the wiki: https://github.com/binary-husky/gpt_academic/wiki/%E9%80%82%E9%85%8DRWKV-Runner

# [Optional Step IV] Make sure that the AVAIL_LLM_MODELS in the config.py configuration file includes the expected models. The currently supported models are as follows (the jittorllms series only supports the docker solution):
AVAIL_LLM_MODELS = ["gpt-3.5-turbo", "api2d-gpt-3.5-turbo", "gpt-4", "api2d-gpt-4", "chatglm", "moss"] # + ["jittorllms_rwkv", "jittorllms_pangualpha", "jittorllms_llama"]

  1. Run
python main.py

Installation Method II: Use Docker

  1. Deploy all the capabilities of the project (this is a large image that includes cuda and latex. However, it is not recommended if your internet speed is slow or your hard disk is small) fullcapacity
# Modify docker-compose.yml, keep scheme 0 and delete the others. Then run:
docker-compose up
  1. ChatGPT+Random Quotes+Wikipedia Summary+Spark and other online models (recommended for most people) basic basiclatex basicaudio
# Modify docker-compose.yml, keep scheme 1 and delete the others. Then run:
docker-compose up

P.S. If you need the Latex plugin feature, please refer to the Wiki. Additionally, you can also use scheme 4 or scheme 0 directly to get the Latex feature.

  1. ChatGPT + ChatGLM2 + MOSS + LLAMA2 + Thousand Questions (Requires familiarity with Nvidia Docker runtime) chatglm
# Modify docker-compose.yml, keep scheme 2 and delete the others. Then run:
docker-compose up

Installation Method III: Other Deployment Methods

  1. One-click run script for Windows. Windows users who are completely unfamiliar with the Python environment can download the one-click run script without local models from the Release section. The script contribution comes from oobabooga.

  2. Use third-party APIs, Azure, etc., Random Quotes, Spark, etc., see the Wiki page.

  3. Pitfall guide for remote deployment on cloud servers. Please visit the cloud server remote deployment wiki

  4. Some new deployment platforms or methods

๊ณ ๊ธ‰ ์‚ฌ์šฉ๋ฒ•

I: ์‚ฌ์šฉ์ž ์ •์˜ ๋ฐ”๋กœ ๊ฐ€๊ธฐ ๋ฒ„ํŠผ ์ถ”๊ฐ€ (ํ•™์ˆ  ๋‹จ์ถ•ํ‚ค)

์ž„์˜์˜ ํ…์ŠคํŠธ ํŽธ์ง‘๊ธฐ๋กœ core_functional.py ํŒŒ์ผ์„ ์—ด๊ณ  ๋‹ค์Œ๊ณผ ๊ฐ™์€ ํ•ญ๋ชฉ์„ ์ถ”๊ฐ€ํ•œ ๋‹ค์Œ ํ”„๋กœ๊ทธ๋žจ์„ ๋‹ค์‹œ ์‹œ์ž‘ํ•˜์‹ญ์‹œ์˜ค. (์ด๋ฏธ ๋ฒ„ํŠผ์ด ์žˆ๋Š” ๊ฒฝ์šฐ์—๋Š” ์ ‘๋‘์‚ฌ์™€ ์ ‘๋ฏธ์‚ฌ๋ฅผ ์‹ค์‹œ๊ฐ„์œผ๋กœ ์ˆ˜์ •ํ•  ์ˆ˜ ์žˆ์œผ๋ฏ€๋กœ ํ”„๋กœ๊ทธ๋žจ์„ ๋‹ค์‹œ ์‹œ์ž‘ํ•  ํ•„์š”๊ฐ€ ์—†์Šต๋‹ˆ๋‹ค.) ์˜ˆ์‹œ:

"์ดˆ๊ธ‰์˜๋ฌธ ๋ฒˆ์—ญ": {
    # ์ ‘๋‘์‚ฌ, ์ž…๋ ฅ ๋‚ด์šฉ ์•ž์— ์ถ”๊ฐ€๋ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด ์š”๊ตฌ ์‚ฌํ•ญ์„ ์„ค๋ช…ํ•˜๋Š” ๋ฐ ์‚ฌ์šฉ๋ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด ๋ฒˆ์—ญ, ์ฝ”๋“œ ์„ค๋ช…, ๊ต์ • ๋“ฑ
    "Prefix": "๋‹ค์Œ ๋‚ด์šฉ์„ ํ•œ๊ตญ์–ด๋กœ ๋ฒˆ์—ญํ•˜๊ณ  ์ „๋ฌธ ์šฉ์–ด์— ๋Œ€ํ•œ ์„ค๋ช…์„ ์ ์šฉํ•œ ๋งˆํฌ๋‹ค์šด ํ‘œ๋ฅผ ์‚ฌ์šฉํ•˜์„ธ์š”:\n\n",

    # ์ ‘๋ฏธ์‚ฌ, ์ž…๋ ฅ ๋‚ด์šฉ ๋’ค์— ์ถ”๊ฐ€๋ฉ๋‹ˆ๋‹ค. ์˜ˆ๋ฅผ ๋“ค์–ด ์ ‘๋‘์‚ฌ์™€ ํ•จ๊ป˜ ์ž…๋ ฅ ๋‚ด์šฉ์„ ๋”ฐ์˜ดํ‘œ๋กœ ๊ฐ์Œ€ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
    "Suffix": "",
},

II: ์‚ฌ์šฉ์ž ์ •์˜ ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ

์›ํ•˜๋Š” ์ž‘์—…์„ ์ˆ˜ํ–‰ํ•˜๊ธฐ ์œ„ํ•ด ๋Šฅ๋ ฅ์žˆ๋Š” ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์ž‘์„ฑํ•˜์„ธ์š”. ์ด ํ”„๋กœ์ ํŠธ์˜ ํ”Œ๋Ÿฌ๊ทธ์ธ ์ž‘์„ฑ ๋ฐ ๋””๋ฒ„๊น…์€ ๋‚œ์ด๋„๊ฐ€ ๋‚ฎ์œผ๋ฉฐ, ์ผ์ •ํ•œ Python ๊ธฐ๋ณธ ์ง€์‹๋งŒ ์žˆ์œผ๋ฉด ์šฐ๋ฆฌ๊ฐ€ ์ œ๊ณตํ•˜๋Š” ํ…œํ”Œ๋ฆฟ์„ ๋ณธ๋”ฐ์„œ ๊ณ ์œ ํ•œ ํ”Œ๋Ÿฌ๊ทธ์ธ ๊ธฐ๋Šฅ์„ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ž์„ธํ•œ ๋‚ด์šฉ์€ ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ ๊ฐ€์ด๋“œ๋ฅผ ์ฐธ์กฐํ•˜์„ธ์š”.

์—…๋ฐ์ดํŠธ

I: ๋‹ค์ด๋‚˜๋ฏน

  1. ๋Œ€ํ™” ์ €์žฅ ๊ธฐ๋Šฅ. ํ”Œ๋Ÿฌ๊ทธ์ธ ์˜์—ญ์—์„œ 'ํ˜„์žฌ ๋Œ€ํ™” ์ €์žฅ'์„ ํ˜ธ์ถœํ•˜์—ฌ ํ˜„์žฌ ๋Œ€ํ™”๋ฅผ ๋ณผ ์ˆ˜ ์žˆ๊ณ , html ํŒŒ์ผ์„ ๋ณต๊ตฌํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ๋˜ํ•œ ํ”Œ๋Ÿฌ๊ทธ์ธ ์˜์—ญ์—์„œ '๋Œ€ํ™” ๊ธฐ๋ก ๋ถˆ๋Ÿฌ์˜ค๊ธฐ'๋ฅผ ํ˜ธ์ถœํ•˜์—ฌ ์ด์ „ ๋Œ€ํ™”๋ฅผ ๋ณต์›ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ํŒ: ํŒŒ์ผ์„ ์ง€์ •ํ•˜์ง€ ์•Š๊ณ  '๋Œ€ํ™” ๊ธฐ๋ก ๋ถˆ๋Ÿฌ์˜ค๊ธฐ'๋ฅผ ๋ฐ”๋กœ ํด๋ฆญํ•˜๋ฉด ์ด์ „ html ๊ธฐ๋ก ์บ์‹œ๋ฅผ ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

  2. โญLatex/Arxiv ๋…ผ๋ฌธ ๋ฒˆ์—ญ ๊ธฐ๋Šฅโญ

    ===>
  3. ๋นˆ ํ„ฐ๋ฏธ๋„ (์ž์—ฐ์–ด ์ž…๋ ฅ์—์„œ ์‚ฌ์šฉ์ž ์˜๋„ ์ดํ•ด + ์ž๋™ ํ”Œ๋Ÿฌ๊ทธ์ธ ํ˜ธ์ถœ)

  • ๋‹จ๊ณ„ 1: "ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ์‚ฌ์šฉํ•˜์—ฌ PDF ๋…ผ๋ฌธ์„ ๋ฒˆ์—ญํ•˜์‹ญ์‹œ์˜ค. ์ฃผ์†Œ๋Š” https://openreview.net/pdf?id=rJl0r3R9KX์ž…๋‹ˆ๋‹ค." ์ž…๋ ฅ
  • ๋‹จ๊ณ„ 2: "๋นˆ ํ„ฐ๋ฏธ๋„" ํด๋ฆญ
  1. ๋ชจ๋“ˆํ™”๋œ ๊ธฐ๋Šฅ ๋””์ž์ธ, ๊ฐ„๋‹จํ•œ ์ธํ„ฐํŽ˜์ด์Šค๋กœ ๊ฐ•๋ ฅํ•œ ๊ธฐ๋Šฅ ์ œ๊ณต

  2. ๋‹ค๋ฅธ ์˜คํ”ˆ ์†Œ์Šค ํ”„๋กœ์ ํŠธ ๋ฒˆ์—ญ

  3. live2d์˜ ์ž‘์€ ๊ธฐ๋Šฅ ์ถ”๊ฐ€ (๊ธฐ๋ณธ ์„ค์ •์€ ๋‹ซํ˜€ ์žˆ์œผ๋ฉฐ, config.py๋ฅผ ์ˆ˜์ •ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.)

  4. OpenAI ์ด๋ฏธ์ง€ ์ƒ์„ฑ

  5. OpenAI ์˜ค๋””์˜ค ๋ถ„์„ ๋ฐ ์š”์•ฝ

  6. Latex ์ „์ฒด ๊ต์ • ์˜ค๋ฅ˜

    ===>
  7. ์–ธ์–ด, ํ…Œ๋งˆ ๋ณ€๊ฒฝ

II: ๋ฒ„์ „:

  • ๋ฒ„์ „ 3.70 (์˜ˆ์ •): AutoGen ํ”Œ๋Ÿฌ๊ทธ์ธ ํ…Œ๋งˆ ๊ฐœ์„  ๋ฐ ๋‹ค๋ฅธ ํ…Œ๋งˆ ํ”Œ๋Ÿฌ๊ทธ์ธ ๋””์ž์ธ
  • ๋ฒ„์ „ 3.60: AutoGen์„ ์ƒˆ๋กœ์šด ์„ธ๋Œ€ ํ”Œ๋Ÿฌ๊ทธ์ธ์˜ ๊ธฐ๋ฐ˜์œผ๋กœ ๋„์ž…
  • ๋ฒ„์ „ 3.57: GLM3, Starfire v3, ๆ–‡ๅฟƒไธ€่จ€ v4 ์ง€์›, ๋กœ์ปฌ ๋ชจ๋ธ์˜ ๋™์‹œ์„ฑ ๋ฒ„๊ทธ ์ˆ˜์ •
  • ๋ฒ„์ „ 3.56: ๋™์ ์œผ๋กœ ๊ธฐ๋ณธ ๊ธฐ๋Šฅ ๋ฒ„ํŠผ ์ถ”๊ฐ€, ์ƒˆ๋กœ์šด ๋ณด๊ณ ์„œ PDF ์š”์•ฝ ํŽ˜์ด์ง€
  • ๋ฒ„์ „ 3.55: ํ”„๋ก ํŠธ ์—”๋“œ ์ธํ„ฐํŽ˜์ด์Šค ๋ฆฌํŒฉํ† ๋ง, ํ™”๋ฉด ๋”ฐ๋ผ๋‹ค๋‹ˆ๋Š” ์œˆ๋„์šฐ ๋ฐ ๋ฉ”๋‰ด ๋ฐ” ๋„์ž…
  • ๋ฒ„์ „ 3.54: ์ƒˆ๋กœ์šด ๋™์  ์ฝ”๋“œ ํ•ด์„๊ธฐ (Code Interpreter) ์ถ”๊ฐ€ (์™„๋ฒฝํ•˜๊ฒŒ ์™„์„ฑ๋˜์ง€ ์•Š์Œ)
  • ๋ฒ„์ „ 3.53: ๋‹ค๋ฅธ ์ธํ„ฐํŽ˜์ด์Šค ํ…Œ๋งˆ ๋™์  ์„ ํƒ ๊ธฐ๋Šฅ ์ถ”๊ฐ€, ์•ˆ์ •์„ฑ ํ–ฅ์ƒ ๋ฐ ๋‹ค์ค‘ ์‚ฌ์šฉ์ž ์ถฉ๋Œ ๋ฌธ์ œ ํ•ด๊ฒฐ
  • ๋ฒ„์ „ 3.50: ์ž์—ฐ์–ด๋กœ ์ด ํ”„๋กœ์ ํŠธ์˜ ๋ชจ๋“  ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์„ ํ˜ธ์ถœํ•˜๋Š” ๊ธฐ๋Šฅ (๋นˆ ํ„ฐ๋ฏธ๋„) ์ถ”๊ฐ€, ํ”Œ๋Ÿฌ๊ทธ์ธ ๋ถ„๋ฅ˜ ์ง€์›, UI ๊ฐœ์„ , ์ƒˆ๋กœ์šด ํ…Œ๋งˆ ์„ค๊ณ„
  • ๋ฒ„์ „ 3.49: Baidu Qianfan ํ”Œ๋žซํผ ๋ฐ ๋ฌธ์‹ฌ์ผ์–ธ ์ง€์›
  • ๋ฒ„์ „ 3.48: Ali DameiYuan Sematic Query, Shanghai AI-Lab Shusheng, Xunfei Starfire ์ง€์›
  • ๋ฒ„์ „ 3.46: ์™„์ „ ์ž๋™ ์šด์ „ ๊ฐ€๋Šฅํ•œ ์‹ค์‹œ๊ฐ„ ์Œ์„ฑ ๋Œ€ํ™” ์ง€์›
  • ๋ฒ„์ „ 3.45: ์‚ฌ์šฉ์ž ์ •์˜ ChatGLM2 fine-tuning ๋ชจ๋ธ ์ง€์›
  • ๋ฒ„์ „ 3.44: Azure ์ •์‹ ์ง€์›, ์ธํ„ฐํŽ˜์ด์Šค์˜ ์‚ฌ์šฉ ํŽธ์˜์„ฑ ๊ฐœ์„ 
  • ๋ฒ„์ „ 3.4: +arxiv ๋…ผ๋ฌธ ๋ฒˆ์—ญ, latex ๋…ผ๋ฌธ ๊ต์ • ๊ธฐ๋Šฅ ์ถ”๊ฐ€
  • ๋ฒ„์ „ 3.3: +์ธํ„ฐ๋„ท ์ •๋ณด ์ข…ํ•ฉ ๊ธฐ๋Šฅ
  • ๋ฒ„์ „ 3.2: ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์ด ๋” ๋งŽ์€ ๋งค๊ฐœ๋ณ€์ˆ˜ ์ธํ„ฐํŽ˜์ด์Šค๋ฅผ ์ง€์›ํ•ฉ๋‹ˆ๋‹ค (๋Œ€ํ™” ์ €์žฅ ๊ธฐ๋Šฅ, ์ž„์˜์˜ ์–ธ์–ด ์ฝ”๋“œ ํ•ด์„ + ์ž„์˜์˜ LLM ์กฐํ•ฉ์„ ๋™์‹œ์— ์š”์ฒญ)
  • ๋ฒ„์ „ 3.1: ์—ฌ๋Ÿฌ GPT ๋ชจ๋ธ์— ๋™์‹œ์— ์งˆ๋ฌธํ•  ์ˆ˜ ์žˆ๋Š” ๊ธฐ๋Šฅ ์ถ”๊ฐ€! api2d ์ง€์›, ์—ฌ๋Ÿฌ ๊ฐœ์˜ apikey ๋ถ€ํ•˜ ๊ท ํ˜• ์กฐ์ • ์ง€์›
  • ๋ฒ„์ „ 3.0: chatglm ๋ฐ ๊ธฐํƒ€ ์†Œ๊ทœ๋ชจ llm ์ง€์›
  • ๋ฒ„์ „ 2.6: ํ”Œ๋Ÿฌ๊ทธ์ธ ๊ตฌ์กฐ๋ฅผ ์žฌ๊ตฌ์„ฑํ•˜์—ฌ ์ƒํ˜ธ ์ž‘์šฉ์„ฑ ํ–ฅ์ƒ, ๋” ๋งŽ์€ ํ”Œ๋Ÿฌ๊ทธ์ธ ์ถ”๊ฐ€
  • ๋ฒ„์ „ 2.5: ์ž๋™ ์—…๋ฐ์ดํŠธ, ์†Œ์Šค ์ฝ”๋“œ ์š”์•ฝ ์ค‘ ํ…์ŠคํŠธ๊ฐ€ ๋„ˆ๋ฌด ๊ธธ๊ณ  ํ† ํฐ์ด ์˜ค๋ฒ„ํ”Œ๋กœ๋˜๋Š” ๋ฌธ์ œ ํ•ด๊ฒฐ
  • ๋ฒ„์ „ 2.4: (1)PDF ์ „์ฒด ๋ฒˆ์—ญ ๊ธฐ๋Šฅ ์ถ”๊ฐ€; (2)์ž…๋ ฅ ์˜์—ญ ์œ„์น˜ ์ „ํ™˜ ๊ธฐ๋Šฅ ์ถ”๊ฐ€; (3)์ˆ˜์ง ๋ ˆ์ด์•„์›ƒ ์˜ต์…˜ ์ถ”๊ฐ€; (4)๋ฉ€ํ‹ฐ ์Šค๋ ˆ๋“œ ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ ์ตœ์ ํ™”
  • ๋ฒ„์ „ 2.3: ๋ฉ€ํ‹ฐ ์Šค๋ ˆ๋“œ ์ƒํ˜ธ ์ž‘์šฉ์„ฑ ๊ฐ•ํ™”
  • ๋ฒ„์ „ 2.2: ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ์˜ ํ•ซ ๋ฆฌ๋กœ๋“œ ์ง€์›
  • ๋ฒ„์ „ 2.1: ์ ‘์„ ์ˆ˜ ์žˆ๋Š” ๋ ˆ์ด์•„์›ƒ
  • ๋ฒ„์ „ 2.0: ๋ชจ๋“ˆํ™” ํ•จ์ˆ˜ ํ”Œ๋Ÿฌ๊ทธ์ธ ๋„์ž…
  • ๋ฒ„์ „ 1.0: ๊ธฐ๋ณธ ๊ธฐ๋Šฅ

GPT Academic ๊ฐœ๋ฐœ์ž QQ ๊ทธ๋ฃน: 610599535

  • ์•Œ๋ ค์ง„ ๋ฌธ์ œ
    • ํŠน์ • ์›น ๋ธŒ๋ผ์šฐ์ € ๋ฒˆ์—ญ ํ”Œ๋Ÿฌ๊ทธ์ธ์ด ์ด ์†Œํ”„ํŠธ์›จ์–ด์˜ ํ”„๋ก ํŠธ์—”๋“œ ์‹คํ–‰์— ๋ฐฉํ•ด๊ฐ€ ๋˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.
    • ๊ณต์‹ Gradio์—๋Š” ํ˜ธํ™˜์„ฑ ๋ฌธ์ œ๊ฐ€ ๋งŽ๊ธฐ ๋•Œ๋ฌธ์— requirement.txt๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ Gradio๋ฅผ ์„ค์น˜ํ•˜์‹ญ์‹œ์˜ค.

III: ํ…Œ๋งˆ

THEME ์˜ต์…˜ (config.py)์„ ์ˆ˜์ •ํ•˜์—ฌ ํ…Œ๋งˆ๋ฅผ ๋ณ€๊ฒฝํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

  1. Chuanhu-Small-and-Beautiful URL

IV: ์ด ํ”„๋กœ์ ํŠธ์˜ ๊ฐœ๋ฐœ ๋ธŒ๋žœ์น˜

  1. master ๋ธŒ๋žœ์น˜: ๋ฉ”์ธ ๋ธŒ๋žœ์น˜, ์•ˆ์ • ๋ฒ„์ „
  2. frontier ๋ธŒ๋žœ์น˜: ๊ฐœ๋ฐœ ๋ธŒ๋žœ์น˜, ํ…Œ์ŠคํŠธ ๋ฒ„์ „

V: ์ฐธ๊ณ  ๋ฐ ํ•™์Šต

์ฝ”๋“œ์—์„œ๋Š” ๋‹ค๋ฅธ ์šฐ์ˆ˜ํ•œ ํ”„๋กœ์ ํŠธ์˜ ๋””์ž์ธ์„ ๋งŽ์ด ์ฐธ๊ณ ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ˆœ์„œ๋Š” ๋ฌธ์ œ ์—†์ด ๋‚˜์—ด๋ฉ๋‹ˆ๋‹ค:

# ๆธ…ๅŽChatGLM2-6B:
https://github.com/THUDM/ChatGLM2-6B

# ๆธ…ๅŽJittorLLMs:
https://github.com/Jittor/JittorLLMs

# ChatPaper:
https://github.com/kaixindelele/ChatPaper

# Edge-GPT:
https://github.com/acheong08/EdgeGPT

# ChuanhuChatGPT:
https://github.com/GaiZhenbiao/ChuanhuChatGPT



# Oobabooga ์› ํด๋ฆญ ์„ค์น˜ ํ”„๋กœ๊ทธ๋žจ:
https://github.com/oobabooga/one-click-installers

# ๋”๋ณด๊ธฐ๏ผš
https://github.com/gradio-app/gradio
https://github.com/fghrsh/live2d_demo