jishnup/embed_kgqa repository overview

Use this docker image for conducting experiments related to MLRC2020-EmbedKGQA⁠.
For the same,

• Use env tag for image without models.
• Use env-models tag for image with models.
  • We have included the models with better performance from the extended study.

A sample command to run on NVIDIA GPU is as follows:

docker run -itd --rm --runtime=nvidia -v /raid/kgdnn/:/raid/kgdnn/ --name embedkgqa__4567 -e NVIDIA_VISIBLE_DEVICES=4,5,6,7  -p 7777:7777 jishnup/embed_kgqa:env

Credits: This image is built upon qts8n/cuda-python⁠, an nvidia/cuda:10.2 based Python 3.7.5 image with some CV libraries.