Sungkyun Chang

@kasireddygariDineshKumarReddy I would try `sudo docker`. If it doesn't work, it is docker installation issue, rather than this repo. Docker installed as rootless can cause this too.

Open your Docker file in the text editor. Insert this line before any apt activities: ``` ENV DEBIAN_FRONTEND=nonintercative ``` https://medium.com/@alexander.tyutin/docker-tips-ubuntu-and-stuck-of-interactive-tools-with-docker-build-1e95383cba96 https://github.com/docker/docker.github.io/issues/13980

It seems something went wrong in the download. However, according to [this](https://github.com/krallin/tini#using-tini), tini.deb is no longer required for Docker >=1.13. Just remove the line of scripts for tini.deb installation.

@guillemcortes I opened the issue as a question #23.

@stdio2016 Hi, Yi-feng. Thank you for questions and reviewing my code. Your implementation looks awesome! I was really waiting for someone else to implement it with PyTorch. I'm on my...

Thanks. Yes actually the training part is same. I have a plan for colab. The g-drive (raw) files are exactly for the purpose of mounting it on colab . Training...

I am wondering if it is possible to install `faiss` (required for constructing search engine) smoothly in colab. I've never tried it yet. It is also an important prerequisite to...

@Mihonarium Thanks for report. Yes, it seems we don't need conda for colab. Just `pip install` works smooth. Installation of faiss-gpu was super smooth too: `!pip install faiss-gpu`. About your...

@Mihonarium Oh it is an expected behavior as I wrote it above. The checkpoint file contains optimzer's states info which is GPU device dependent. So, if you wanna continue train...

@Mihonarium About training from scratch error: First, for P100 GPU, I recommend ``` BSZ: TR_BATCH_SZ : 320 # Training batch size N must be EVEN number. TR_N_ANCHOR : 160 ```...