ModelScope Serving

Use the following command to easily start ModelScope Serving

git clone https://github.com/alphasnow/modelscope-serving.git
cd modelscope-serving
docker-compose up -f docker-compose.yml -d

Example

# request
curl 'http://127.0.0.1:8800/api/predict/image-cartoon' \
-H 'Content-Type:application/json' \
-X POST \
-d '{"image":"TdbnJ6cHh9mV09O5qu+z6Sv3j06vPvJvcePHqdrLa...","algo_type":"anime"}'
# response
{"image":"Yc+a9/V3bHXmnZSu26oXaBa66pbLpbd4f7+8dHxj370o..."}

Features

• Automatic Speech Recognition /asr/paraformer
• Text To Speech /tts/emotion
• Photo to Cartoon /image-cartoon
• Image segmentation /image-segment/person /image-segment/product
• Optical Character Recognition /image-ocr/general
• Word Segment /word-segmentation
• Body Detection /image-body/2d-keypoint

Dependencies

• docker
• docker-compose
• modelscope
• python
• fastapi
• opencv
• uvicorn

License

This project is licensed under the Apache License (Version 2.0).