Codegen

Salesforce codegen with web server

Fork from https://github.com/salesforce/CodeGen

Add web server support

Support GPT-J Now

Online demo

https://gitclone.com/aiit/chat/

Visual Studio Code Extension

https://github.com/git-cloner/codegeeker

usage

Clone code

git clone https://gitclone.com/github.com/git-cloner/codegen

Install

1.install Nvidia Graphics Card and Driver

2.install conda

3.init vitual runtime

conda create -n codegen python=3.8
conda activate codegen
pip install pillow -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install torch torchvision torchaudio -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install transformers==4.25.1 -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install accelerate -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install aiohttp==3.8.3 -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install aiohttp_cors==0.7.0 -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
#pip install huggingface for gpt-neo
pip install datasets -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install gradio -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com
pip install sentencepiece==0.1.91 -i http://pypi.douban.com/simple --trusted-host=pypi.douban.com

Download model

wget -P checkpoints https://storage.googleapis.com/sfr-codegen-research/checkpoints/codegen-350M-nl.tar.gz && tar -xvf checkpoints/codegen-350M-nl.tar.gz -C checkpoints/
wget -P checkpoints https://storage.googleapis.com/sfr-codegen-research/checkpoints/codegen-350M-multi.tar.gz && tar -xvf checkpoints/codegen-350M-multi.tar.gz -C checkpoints/
wget -P checkpoints https://storage.googleapis.com/sfr-codegen-research/checkpoints/codegen-350M-mono.tar.gz && tar -xvf checkpoints/codegen-350M-mono.tar.gz -C checkpoints/

run as web server

conda activate codegen
python codegen.py

test

post: http://127.0.0.1:5000/codegen or direct post to :https://gitclone.com/aiit/codegen

Use Salesforce codegen

input params: {"context":"def hello_world():","maxlength":128}

return params: { "result": "\n print("Hello world")\n hello_world()\n ", "time": 1.9620850095525384 }

Use GPT-J

input params: {"context":"写一个python版的数组排序","maxlength":128}