# Infrastructure & Execution Environment

## Training & Experimentation

### Data Ingestion & Preprocessing

- [Docker Containerization & Execution Environment](https://github.com/dessa-oss/deepfake-detection/tree/master/custom_docker_image) 
- [Experiment Config (job.config.yaml)](https://github.com/dessa-oss/deepfake-detection/blob/master/job.config.yaml)

### Atlas Experiment Manager

### External Tools

- NVIDIA-docker
- ffmpeg

### Model Definition

- [model.py](https://github.com/dessa-oss/deepfake-detection/blob/master/model.py)

### Pretrained Model

- ResNet18 (ImageNet)

### Training Inference

- [main.py](https://github.com/dessa-oss/deepfake-detection/blob/master/main.py)
- [train.py](https://github.com/dessa-oss/deepfake-detection/blob/master/train.py)
- [Hyperparameter Search (hparams_search.py)](https://github.com/dessa-oss/deepfake-detection/blob/master/hparams_search.py)
- [faceforensics_download.py](https://github.com/dessa-oss/deepfake-detection/blob/master/faceforensics_download.py)
- [download_T.sh](https://github.com/dessa-oss/deepfake-detection/blob/master/download_T.sh)
- [restructure_data.sh](https://github.com/dessa-oss/deepfake-detection/blob/master/restructure_data.sh)
- [data_loader.py](https://github.com/dessa-oss/deepfake-detection/blob/master/data_loader.py)

### Datasets

- runs
- prepares
- loads
- input
- initializes
- triggers
- optimizes
- configures
- schedules
- supports
- executes via
- runs