diff --git a/README.md b/README.md
index a91680a084ba696cbde84a73c1a20b003c495976..84c1664c21164a8130e77990d5f947bdbb5cdda7 100644
--- a/README.md
+++ b/README.md
@@ -8,7 +8,7 @@ This project aims to adopt the GAN-based architectures,  which original proposed
 - Python 3
 - CPU or NVIDIA GPU + CUDA CuDNN
 
-### Installation
+### Installation 
 - Clone this repo:
 ```bash
 git clone -b master https://gitlab.version.fz-juelich.de/gong1/video_prediction_savp.git
@@ -21,6 +21,17 @@ cd Video_Prediction_SAVP
 pip install -r requirements.txt
 ```
 
+###Set-up on JUWELS
+
+- Set up env and install packages
+
+```bash
+cd env_setup
+./create_env.sh <USER_FOLDER>
+```
+
+
+
 ### Miscellaneous installation considerations
 - In python >= 3.6, make sure to add the root directory to the PYTHONPATH`, e.g. `export PYTHONPATH=path/to/video_prediction_savp`.
 - For the best speed and experimental results, we recommend using cudnn version 7.3.0.29 and any tensorflow version >= 1.9 and <= 1.12. The final training loss is worse when using cudnn versions 7.3.1.20 or 7.4.1.5, compared to when using versions 7.3.0.29 and below.
@@ -38,7 +49,6 @@ bash data/download_and_preprocess_dataset_era5.sh --data era5 --input_dir /split
 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
 ```
 
-
 ### Model Evaluation
 
 ![Groud Truth](/results_test_samples/era5_size_64_64_3_norm_dup/ours_savp/Sample_Batch_id_0_Sample_1.mp4)