update slurm sbatch examples

gagandaroach · Feb 2, 2021 · 14c0608 · 14c0608
1 parent 8e985a1
commit 14c0608
Show file tree

Hide file tree

Showing 10 changed files with 80 additions and 157 deletions.
diff --git a/slurm/README.md b/slurm/README.md
@@ -10,4 +10,4 @@ To execute these scripts, use sbatch on the management nodes.
     $ sbatch train_1gpu_t4.sh
 ```
 
-**Note:** Read **srun** and **sbatch** manual pages for detailed run parameters.
+**Note:** Read **srun** and **sbatch** manual pages for detailed run parameters. `$ man sbatch`. 
diff --git a/slurm/cpu_basic.sh b/slurm/cpu_basic.sh
@@ -0,0 +1,18 @@
+#!/bin/bash
+
+#SBATCH --job-name="Sbatch Example"
+#SBATCH --output=cpu_job_%j.out
+#SBATCH --mail-type=ALL
+#SBATCH [email protected]
+#SBATCH --partition=teaching
+#SBATCH --nodes=1
+#SBATCH --cpus-per-task=2
+#SBATCH --mem=2GB
+
+## SCRIPT START
+
+srun echo "Hello from the executing node!"
+srun hostname
+srun python --version
+
+## SCRIPT END
diff --git a/slurm/dgx_singularity.sh b/slurm/dgx_singularity.sh
@@ -0,0 +1,21 @@
+#!/bin/bash
+
+#SBATCH --job-name="DGX 8 GPU"
+#SBATCH --output=train_%N_%j.out
+#SBATCH --mail-type=ALL
+#SBATCH [email protected]
+#SBATCH --partition=dgx
+#SBATCH --nodes=1
+#SBATCH --gres=gpu:v100:8
+#SBATCH --cpus-per-gpu=8
+
+SCRIPT_NAME="Rosie DGX Script"
+CONTAINER="/data/containers/msoe-tensorflow.sif"
+SCRIPT_PATH=""
+SCRIPT_ARGS=""
+
+## SCRIPT
+echo "SBATCH SCRIPT: ${SCRIPT_NAME}"
+srun hostname; pwd; date;
+srun singularity exec --nv -B /data:/data ${CONTAINER} python3 ${SCRIPT_PATH} ${SCRIPT_ARGS}
+echo "END: " $SCRIPT_NAME
diff --git a/slurm/sbatch_example.sh b/slurm/sbatch_example.sh
diff --git a/slurm/schedule_python_script.sh b/slurm/schedule_python_script.sh
diff --git a/slurm/t4_gpu.sh b/slurm/t4_gpu.sh
@@ -0,0 +1,19 @@
+#!/bin/bash
+
+#SBATCH --job-name="Sbatch Example"
+#SBATCH --output=job_%j.out
+#SBATCH --mail-type=ALL
+#SBATCH [email protected]
+#SBATCH --partition=teaching
+#SBATCH --nodes=1
+#SBATCH --gres=gpu:t4:1
+#SBATCH --cpus-per-gpu=4
+
+## SCRIPT START
+
+srun echo "Hello from the executing node!"
+srun hostname
+srun python --version
+srun nvidia-smi
+
+## SCRIPT END
diff --git a/slurm/t4_singularity.sh b/slurm/t4_singularity.sh
@@ -0,0 +1,21 @@
+#!/bin/bash
+
+#SBATCH --job-name="Sbatch Example"
+#SBATCH --output=job_%j.out
+#SBATCH --mail-type=ALL
+#SBATCH [email protected]
+#SBATCH --partition=teaching
+#SBATCH --nodes=1
+#SBATCH --gres=gpu:t4:1
+#SBATCH --cpus-per-gpu=4
+
+SCRIPT_NAME="Rosie Job Script"
+CONTAINER="/data/containers/msoe-tensorflow.sif"
+SCRIPT_PATH=""
+SCRIPT_ARGS=""
+
+## SCRIPT
+echo "SBATCH SCRIPT: ${SCRIPT_NAME}"
+srun hostname; pwd; date;
+srun singularity exec --nv -B /data:/data ${CONTAINER} python3 ${SCRIPT_PATH} ${SCRIPT_ARGS}
+echo "END: " $SCRIPT_NAME
diff --git a/slurm/train_1gpu_t4.sh b/slurm/train_1gpu_t4.sh
diff --git a/slurm/train_2gpu_t4.sh b/slurm/train_2gpu_t4.sh
diff --git a/slurm/train_8gpu_dgx.sh b/slurm/train_8gpu_dgx.sh