Upload New File

manual/cluster_job_GPU_processing_template.job

+#!/usr/bin/zsh
+
+##############################################
+##### Batch script for the MRCNN processing ####
+##############################################
+
+#### CREATE SBATCH ENTRIES ####
+#### Paths and parameters must be adapted accordingly. 
+
+#### job name 
+#SBATCH --job-name=<JobName>
+
+#### Path and name of the output file of the job execution 
+#SBATCH --output=/home/<UserID>/.../<JobOutputFolderName>/%x_%J_output.txt
+
+#### Job runtime  
+#SBATCH --time=0-00:00:00
+
+#### Memory requirement per GPU . 
+#### For example: if value is 5GB --> --mem-per-gpu=5G 
+#SBATCH --mem-per-gpu=5G
+
+#### E-mail address 
+#SBATCH --mail-user=<EmailAdress>
+
+#### E-mails to be received 
+#SBATCH --mail-type=ALL
+
+#### Number of tasks to be performed 
+#SBATCH --ntasks=1
+
+#### Number of GPUs required per node 
+#SBATCH --gres=gpu:1
+
+#### Definition of the job array starting at 0. ###
+#### This parameter is only required if you want to perform several jobs in parallel 
+#### from one job script, e.g. processing one testing image set with several MRCNN models (epochs) 
+#### In this example we process one testing image set with 10 MRCNN models (= 10 epochs).
+#### Thus, we will run 10 jobs in parallel from one job script --> array=0-9 
+#SBATCH --array=0-9
+
+#### CREATE TERMINAL ENTRIES ####
+#### Paths and parameters must be adapted accordingly 
+
+#### Definition of the job parameter, which is varied 
+#### if several jobs are executed in parallel from one job script. 
+#### This job parameter is only required if you have specified the #SBATCH parameter --array above. 
+#### In this example, we process one testing image set with 10 MRCNN models.
+#### Thus, we will run 10 jobs in parallel from one job script: 
+#### the parameter model corresponds to the model of the current processing, 
+#### which is varied for each job.
+model="$SLURM_ARRAY_TASK_ID"
+
+#### Loading the Cuda module 
+module load cuda/10.0
+
+#### Export path in which Anaconda is located 
+export PATH=$PATH:/home/<UserID>/anaconda3/bin
+
+#### Activate environment 
+source activate env_mrcnn_gpu
+
+#### Navigate to the path where the droplet.py script is located
+cd /home/<UserID>/.../samples/droplet/
+
+#### Run the process_automated_droplet.py script. 
+#### These are the required processing parameters to be specified 
+#### with additional parameters required for the execution of parallel jobs from one job script. 
+#### In this example, we process one testing image set with 10 MRCNN models.
+#### Thus, 10 jobs are executed in parallel (#SBATCH --array=0-9).
+#### In each job the job parameter model is varied, starting with 0 and ending with 9.
+#### The model names are model_00 to model_09.
+#### First, we specify the processing parameter weights_name (--weights_name=model_0"$model").
+#### Moreover, we specify output folder and Excel output file names 
+#### defined by the processing parameters save_path and name_result_file, since we need 10 of them.  
+#### Optional processing parameters can be found below. 
+#### Description/default settings of all processing parameters see manual. 
+python process_automated_droplet.py  --dataset_path=<InputFolderName> --save_path=<OutputFolderName>_0"$model" --name_result_file=<ExcelFileName>_0"$model" --weights_path=<WeightsFolderName> --weights_name=model_0"$model" --file_format=<FileFormat> --device=<Boolean> --pixelsize=<Double> --image_max=<Integer> 
+
+#### Optional processing parameters: 
+#### --masks
+#### --save_nth_image
+#### --image_crop
+#### --images_gpu
+#### --confidence
+#### --detect_reflections
+#### --detect_oval_droplets
+#### --min_aspect_ratio     
+#### --detect_adhesive_droplets
+#### --save_coordinates
+#### --min_velocity
+#### --min_size_diff
+#### --n_images_compared
+#### --n_adhesive_high
+#### --n_adhesive_low
+#### --low_distance_threshold
+#### --edge_tolerance
+#### --contrast
\ No newline at end of file