From 8cb9dea3a349ed42db5b7850f08ca712c6a67c1d Mon Sep 17 00:00:00 2001
From: Jan Francu <jan.francu@gmail.com>
Date: Mon, 3 May 2021 23:00:10 +0200
Subject: [PATCH] Added acknowledgement.

---
 scripts/experiments_tabular/dagmm_run.sh |  5 +++--
 src/models/dagmm.jl                      | 10 ++++++++++
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/scripts/experiments_tabular/dagmm_run.sh b/scripts/experiments_tabular/dagmm_run.sh
index 4cbef7cd..227d84d5 100644
--- a/scripts/experiments_tabular/dagmm_run.sh
+++ b/scripts/experiments_tabular/dagmm_run.sh
@@ -1,7 +1,8 @@
 #!/bin/bash
+#SBATCH --partition=cpu
 #SBATCH --time=24:00:00
-#SBATCH --nodes=1 --ntasks-per-node=4 --cpus-per-task=1
-#SBATCH --mem=24G
+#SBATCH --nodes=1 --cpus-per-task=1
+#SBATCH --mem=16G
 #SBATCH --qos==collaborator
 
 MAX_SEED=$1
diff --git a/src/models/dagmm.jl b/src/models/dagmm.jl
index eaf9f31a..ba45a31e 100644
--- a/src/models/dagmm.jl
+++ b/src/models/dagmm.jl
@@ -1,3 +1,13 @@
+"""
+    Implements all needed parts for constructing DAGMM model
+        (https://sites.cs.ucsb.edu/~bzong/doc/iclr18-dagmm.pdf) for detecting anomalies.
+    Code is inspired by multiple reimplementations
+    - https://github.com/danieltan07/dagmm
+    - https://github.com/tnakae/DAGMM/
+    - https://github.com/Newcomer520/tf-dagmm
+    - https://github.com/mperezcarrasco/PyTorch-DAGMM
+"""
+
 using Flux: mse, softmax, unsqueeze, stack
 using LinearAlgebra
 using StatsBase