Skip to content
Snippets Groups Projects
Commit 2bd157e5 authored by Bing Gong's avatar Bing Gong
Browse files

update env, readme

parent d6be3e29
Branches
Tags
No related merge requests found
......@@ -12,7 +12,7 @@ This project aims to adopt the GAN-based architectures, which original proposed
- Clone this repo:
```bash
git clone -b master https://gitlab.version.fz-juelich.de/gong1/video_prediction_savp.git
cd Video_Prediction_SAVP
cd video_prediction_savp
```
- Install TensorFlow >= 1.9 and dependencies from http://tensorflow.org/
- Install other dependencies
......@@ -36,16 +36,48 @@ cd env_setup
./create_env.sh <USER_FOLDER>
```
## Workflow by steps
## Data Extraction
## Data Preprocessing
```bash
cd workflow_video_prediction/DataExtraction
python mpi_stager_v2.py
```
### Download data
- Download the ERA5 data (.hkl) from the output of DataPreprocess in the [Workflow project](https://gitlab.version.fz-juelich.de/gong1/workflow_parallel_frame_prediction)
```bash
bash data/download_and_preprocess_dataset_era5.sh --data era5 --input_dir /splits --output_dir data/era5
bash data/download_and_preprocess_dataset_era5.sh --data era5 --input_dir /splits --output_dir <data/era5>
```
## Training
## Postprocessing
### Prediction Generator
### Model Evaluation
## End-to-End run the entire workflow
```bash
cd bash
./workflow_era5.sh
```
### Download data
- Download the ERA5 data (.hkl) from the output of DataPreprocess in the [Workflow project](https://gitlab.version.fz-juelich.de/gong1/workflow_parallel_frame_prediction)
### Model Training for ERA5 data
```python
python scripts/train_v2.py --input_dir data/era5 --dataset era5 --model savp --model_hparams_dict hparams/kth/ours_savp/model_hparams.json --output_dir logs/era5/ours_savp
......
......@@ -6,7 +6,7 @@ if [[ ! -n "$1" ]]; then
fi
FOLDER_NAME=$1
WORKING_DIR=/p/project/deepacf/deeprain/${FOLDER_NAME}/Video_Prediction_SAVP
WORKING_DIR=/p/project/deepacf/deeprain/${FOLDER_NAME}/video_prediction_savp
ENV_DIR=${WORKING_DIR}/env_setup
source ${ENV_DIR}/modules.sh
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment