ivaniscoding / resullme Goto Github PK

Enhance your résumé with Large Language Models

Home Page: https://resullme.streamlit.app/

License: MIT License

Python 99.13% Dockerfile 0.87%

generative-ai json-resume large-language-models latex-resume latex-resume-template resume resume-template streamlit collaborate learn

resullme's Introduction

ResuLLMe

👉 Check some sample résumés generated by ResuLLMe (1, 2, 3)

🚀 Concept

ResuLLMe is a prototype that uses Large Language Models (LLMs) to tailor résumés. It's goal is to enhance résumés to help candidates avoid common mistakes that occur while applying for jobs. It is like a smart career advisor to check your résumé.

You can use ResuLLMe live at https://resullme.streamlit.app/.

🛠 How It Works

ResuLLMe receives your previous CV as a PDF or Word Document. Then, it uses LLMs to:

Improve the résumé following published résumé guidelines by well-reputed schools
Convert the résumés to a JSON Resume format
Render the JSON resume using LaTeX to generate a new PDF of the enhanced resume

🏃 Running

To run ResuLLMe locally, the simplest way is to use Docker:

docker-compose up -d

This will make the app avaialable at https://localhost:8501/

🪄 Installation Instructions for Running Natively

To run the app without Docker, you will need to install two things for the app to work. The first item is to install the Python dependencies:

pip install -r requirements.txt

The second item is to install the LaTeX packages:

xargs sudo apt install -y < packages.txt

Lastly, to run ResuLLMe locally, execute:

streamlit run src/Main.py

🤲 Contributing

ResuLLMe is an open source project.

If you want to contribute, open a Pull requests. All contributions are welcome, but some that would particularly be useful to the community are:

Fixes in existing LaTeX templates
Adding new LaTeX templates
Improved prompts
Support for other LLMs (e.g. Bard, Claude, LLaMA)

resullme's People

Contributors

Stargazers

Watchers

Forkers

poturevanth allowebs lokendarjangid byted4nce hriaz17 elsebaiymohamed ninjasstudio dzungquoc luceinrock albinjosep rmasiniexpert darklord-5 masshat kartheekmedathati jkcheng pgruchot deceptivescorpion jasonnathan jackman337 acse-hy23 tblakex01 krishna7085 huyle82us valeriylo crawford30 alil2000 okoliechykwuka awitkowski0 78ventures yashc1308 pangwa bianyuanop coldn00dles jonariley anisgodev databooth catspunch sumerudataanalaytics drroad tysoncung kakkartushar1 dylanrubini ariafyy iamashwin99 benhurjoel melochineer midhunsankar23 followupper charlysan deyh2020 s-dawn anishekkamal cedrico5000 karankhatavkar yaswanthraj-sp majid-nisar hbcbh1999 digambar9354 dion11235 codegrande unaisshemim mr-haseeb devlincashman yscodehub supersonic13 tejaswi-mahanthi erikwco vanshgulati16 alberto-codes willassaly chinpeerapat pasupuletinarasimha999 kuntalojha davidshan40 suyogb uzzielperez nymeria-direwolf jsoncm

resullme's Issues

Allow for users to specify which OPENAI model

and/or just set the default to gpt-4o

Creator of JSON Resume here, I've been building out AI tools for the project and gpt-4o is far superior. (cheap as chips too)

I don't know how I didn't see this project, it is super awesome, lots of cool ideas. (added it to official projects page)

App not working

Add Docker support

It will help people run the app without LaTeX issues

Add template to Overleaf

To help people manually edit their resumes

Add publication section

A lot of academic CVs contain a list of publications, could we add support for a publication section?

udpate documentation | README.md

before usign "" docker-compose up -d ""

user have to install docker-compose ::

"" sudo apt install docker-compose ""
into the system

Fix Link Issue on Alta Resume

It adds https twice to the link in resume

Didn't work

resume.pdf
It gave a sample.pdf -Didn't work

No output

Many thanks for your work. I am wondering why I cannot get the results after feeding the resume into the app, it shows tailoring and running but nothing happened later.

Add support for other LLMs

Investigate adding support for:

Bard
Claude
Llama

It woud also be nice to host the Llama version on HuggingFace such that people don't need an API token

Getting an AttributeError: module 'openai' has no attribute 'error' when running from docker

Full stack is:

File "/usr/local/lib/python3.10/dist-packages/streamlit/runtime/scriptrunner/script_runner.py", line 565, in _run_script
exec(code, module.dict)
File "/app/src/Main.py", line 104, in
except openai.error.RateLimitError as e:

I assume it is related to the version of openai in the requirements file:

openai
langchain
pdfminer-six
python-docx
streamlit==1.21.0
Jinja2
docx2txt
streamlit-ext==0.1.7
stqdm==0.0.5

Since it is not hard coding the version, it may be using a newer version that is being affected by the ratelimiter.

Very long given name
Very long university name
Lots of skills in a category

The templates that are most affected by it are:

Deedy
Plush
Simple

-result pdf
resume.pdf