From d98af5a7277a2a8bc06c6aea8e1acb2fda2422f3 Mon Sep 17 00:00:00 2001 From: Bing Gong <gongbing1112@gmail.com> Date: Mon, 2 Mar 2020 14:26:45 +0100 Subject: [PATCH] add HPC_scripts --- HPC_scripts/DataPreprocess.sh | 29 +++++++++++++++++++++++++++++ HPC_scripts/DataPreprocess_dev.sh | 29 +++++++++++++++++++++++++++++ HPC_scripts/generate_era5.sh | 26 ++++++++++++++++++++++++++ HPC_scripts/preprocessData_dev.sh | 23 +++++++++++++++++++++++ HPC_scripts/train_era5.sh | 28 ++++++++++++++++++++++++++++ 5 files changed, 135 insertions(+) create mode 100755 HPC_scripts/DataPreprocess.sh create mode 100755 HPC_scripts/DataPreprocess_dev.sh create mode 100755 HPC_scripts/generate_era5.sh create mode 100755 HPC_scripts/preprocessData_dev.sh create mode 100755 HPC_scripts/train_era5.sh diff --git a/HPC_scripts/DataPreprocess.sh b/HPC_scripts/DataPreprocess.sh new file mode 100755 index 00000000..873dbf29 --- /dev/null +++ b/HPC_scripts/DataPreprocess.sh @@ -0,0 +1,29 @@ +#!/bin/bash -x +#SBATCH --account=deepacf +#SBATCH --nodes=1 +#SBATCH --ntasks=1 +##SBATCH --ntasks-per-node=1 +#SBATCH --cpus-per-task=1 +#SBATCH --output=DataPreprocess-out.%j +#SBATCH --error=DataPreprocess-err.%j +#SBATCH --time=20:00:00 +#SBATCH --gres=gpu:1 +#SBATCH --partition=gpus +#SBATCH --mail-type=ALL +#SBATCH --mail-user=b.gong@fz-juelich.de +##jutil env activate -p cjjsc42 + +module --force purge +module /usr/local/software/jureca/OtherStages +module load Stages/2019a +module load GCCcore/.8.3.0 +module load mpi4py/3.0.1-Python-3.6.8 +module load h5py/2.9.0-serial-Python-3.6.8 +module load TensorFlow/1.13.1-GPU-Python-3.6.8 +module load cuDNN/7.5.1.10-CUDA-10.1.105 + + +srun python video_prediction/datasets/era5_dataset_v2.py /p/scratch/deepacf/bing/processData_size_64_64_3_2016/splits/ ./data/era5_size_64_64_3_norm_2016 +: +#srun python scripts/generate_transfer_learning.py --input_dir data/era5_size_64_64_1_v2 --dataset_hparams sequence_length=20 --checkpoint pretrained_models/kth/ours_savp --mode test --results_dir results_test_samples/era5_size_64_64_1_v2_pretrained --dataset era5 +#srun python scripts/train.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 diff --git a/HPC_scripts/DataPreprocess_dev.sh b/HPC_scripts/DataPreprocess_dev.sh new file mode 100755 index 00000000..5aecf65b --- /dev/null +++ b/HPC_scripts/DataPreprocess_dev.sh @@ -0,0 +1,29 @@ +#!/bin/bash -x +#SBATCH --account=deepacf +#SBATCH --nodes=1 +#SBATCH --ntasks=1 +##SBATCH --ntasks-per-node=1 +#SBATCH --cpus-per-task=1 +#SBATCH --output=DataPreprocess-out.%j +#SBATCH --error=DataPreprocess-err.%j +#SBATCH --time=00:20:00 +#SBATCH --gres=gpu:1 +#SBATCH --partition=develgpus +#SBATCH --mail-type=ALL +#SBATCH --mail-user=b.gong@fz-juelich.de +##jutil env activate -p cjjsc42 + +module --force purge +module /usr/local/software/jureca/OtherStages +module load Stages/2019a +module load GCCcore/.8.3.0 +module load mpi4py/3.0.1-Python-3.6.8 +module load h5py/2.9.0-serial-Python-3.6.8 +module load TensorFlow/1.13.1-GPU-Python-3.6.8 +module load cuDNN/7.5.1.10-CUDA-10.1.105 + + +srun python video_prediction/datasets/era5_dataset_v2.py /p/scratch/deepacf/bing/processData_size_64_64_3_dup/splits/ ./data/era5_size_64_64_3_norm_dup +: +#srun python scripts/generate_transfer_learning.py --input_dir data/era5_size_64_64_1_v2 --dataset_hparams sequence_length=20 --checkpoint pretrained_models/kth/ours_savp --mode test --results_dir results_test_samples/era5_size_64_64_1_v2_pretrained --dataset era5 +#srun python scripts/train.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 diff --git a/HPC_scripts/generate_era5.sh b/HPC_scripts/generate_era5.sh new file mode 100755 index 00000000..e0481b47 --- /dev/null +++ b/HPC_scripts/generate_era5.sh @@ -0,0 +1,26 @@ +#!/bin/bash -x +#SBATCH --account=deepacf +#SBATCH --nodes=1 +#SBATCH --ntasks=1 +##SBATCH --ntasks-per-node=1 +#SBATCH --cpus-per-task=1 +#SBATCH --output=generate_era5-out.%j +#SBATCH --error=generate_era5-err.%j +#SBATCH --time=00:20:00 +#SBATCH --gres=gpu:1 +#SBATCH --partition=develgpus +#SBATCH --mail-type=ALL +#SBATCH --mail-user=b.gong@fz-juelich.de +##jutil env activate -p cjjsc42 + + + +module load GCC/8.3.0 +module load ParaStationMPI/5.2.2-1 +module load TensorFlow/1.13.1-GPU-Python-3.6.8 +module load netcdf4-python/1.5.0.1-Python-3.6.8 +module load h5py/2.9.0-Python-3.6.8 +source mandarine/bin/activate + +python scripts/generate_transfer_learning_finetune.py --input_dir data/era5_size_64_64_3_norm_dup --dataset_hparams sequence_length=20 --checkpoint logs/era5_64_64_3_norm_2016/ours_savp --mode test --results_dir results_test_samples/era5_size_64_64_3_norm_2016 --batch_size 4 --dataset era5 +#srun python scripts/train.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 diff --git a/HPC_scripts/preprocessData_dev.sh b/HPC_scripts/preprocessData_dev.sh new file mode 100755 index 00000000..a8828bcc --- /dev/null +++ b/HPC_scripts/preprocessData_dev.sh @@ -0,0 +1,23 @@ +#!/bin/bash -x +#SBATCH --account=deepacf +#SBATCH --nodes=1 +#SBATCH --ntasks=1 +##SBATCH --ntasks-per-node=1 +#SBATCH --cpus-per-task=1 +#SBATCH --output=preprocess-out.%j +#SBATCH --error=preprocess-err.%j +#SBATCH --time=20:20:00 +#SBATCH --partition=batch +#SBATCH --mail-type=ALL +#SBATCH --mail-user=b.gong@fz-juelich.de +##jutil env activate -p cjjsc42 + +module --force purge +module /usr/local/software/jureca/OtherStages +module load Stages/2019a +module load GCCcore/.8.3.0 +module load mpi4py/3.0.1-Python-3.6.8 +module load h5py/2.9.0-serial-Python-3.6.8 +module load TensorFlow/1.13.1-GPU-Python-3.6.8 + +srun bash data/download_and_preprocess_dataset_era5.sh --data era5 --input_dir /p/scratch/deepacf/bing/processData_size_64_64_3/splits --output_dir data/era5_size_64_64_3/ours_savp diff --git a/HPC_scripts/train_era5.sh b/HPC_scripts/train_era5.sh new file mode 100755 index 00000000..3093bdb2 --- /dev/null +++ b/HPC_scripts/train_era5.sh @@ -0,0 +1,28 @@ +#!/bin/bash -x +#SBATCH --account=deepacf +#SBATCH --nodes=1 +#SBATCH --ntasks=1 +##SBATCH --ntasks-per-node=1 +#SBATCH --cpus-per-task=1 +#SBATCH --output=train_era5-out.%j +#SBATCH --error=train_era5-err.%j +#SBATCH --time=23:20:00 +#SBATCH --gres=gpu:1 +#SBATCH --partition=gpus +#SBATCH --mail-type=ALL +#SBATCH --mail-user=b.gong@fz-juelich.de +##jutil env activate -p cjjsc42 + +module --force purge +module /usr/local/software/jureca/OtherStages +module load Stages/2019a +module load GCCcore/.8.3.0 +module load mpi4py/3.0.1-Python-3.6.8 +module load h5py/2.9.0-serial-Python-3.6.8 +module load TensorFlow/1.13.1-GPU-Python-3.6.8 +module load cuDNN/7.5.1.10-CUDA-10.1.105 + +#srun python scripts/train.py --input_dir data/kth --dataset kth --model savp --model_hparams_dict hparams/kth/ours_savp/model_hparams.json --output_dir logs/kth/ours_savp + +srun python scripts/train_v2.py --input_dir data/era5_size_128_128_1 --dataset era5 --model savp --model_hparams_dict hparams/kth/ours_savp/model_hparams.json --output_dir logs/era5_128_128_1/ours_savp +#srun python scripts/train.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 -- GitLab