7th place solution to Topcoder's "Predict the torque from the bolt tightening sound" challenge

Competition website

In this problem, one is given an audio file containing the sound made while bolting steel material on a construction site. The code should read the audio file and predict the torque value of it.

Instructions to run the code

System Requirements

The following system requirements should be satisfied:

• OS: Ubuntu 16.04
• Python: 3.6
• CUDA: 10.1
• cudnn: 7
• Docker

Environment Setup

1. Build a docker image: docker build -t topcoder .
2. Start a docker container:

docker run --rm --gpus all --shm-size 16G \
-v /path/to/train/audios:/data/input/train \
-v /path/to/train/labels:/data/gt/train \
-v /path/to/save/models:/tmp \
-v /path/to/test/audios:/data/input/pred \
-v /path/to/test/predictions:/data/output/pred \
-it topcoder

Build the Model

1. Use /code/train.sh to train the model
2. Use /code/test.sh to make the inference