From 85308196d3d85439af008f093eba1875eee4155e Mon Sep 17 00:00:00 2001
From: AntreasAntoniou <antrikohs@gmail.com>
Date: Fri, 9 Feb 2018 22:01:22 +0000
Subject: [PATCH] Updated tqdm counting issue, added script for setting up
 environment on head-node

---
 cifar100_network_trainer.py       |  2 +-
 cifar10_network_trainer.py        |  2 +-
 emnist_network_trainer.py         |  2 +-
 gpu_cluster_environment_script.sh | 20 ++++++++++++++++++++
 msd10_network_trainer.py          |  2 +-
 msd25_network_trainer.py          |  2 +-
 6 files changed, 25 insertions(+), 5 deletions(-)
 create mode 100644 gpu_cluster_environment_script.sh

diff --git a/cifar100_network_trainer.py b/cifar100_network_trainer.py
index e5cc4de..54de53c 100644
--- a/cifar100_network_trainer.py
+++ b/cifar100_network_trainer.py
@@ -83,7 +83,7 @@ with tf.Session() as sess:
                                                    continue_from_epoch))  # restore previous graph to continue operations
 
     best_val_accuracy = 0.
-    with tqdm.tqdm(total=epochs) as epoch_pbar:
+    with tqdm.tqdm(total=epochs - start_epoch) as epoch_pbar:
         for e in range(start_epoch, epochs):
             total_c_loss = 0.
             total_accuracy = 0.
diff --git a/cifar10_network_trainer.py b/cifar10_network_trainer.py
index d456c9c..6cf61f3 100644
--- a/cifar10_network_trainer.py
+++ b/cifar10_network_trainer.py
@@ -83,7 +83,7 @@ with tf.Session() as sess:
                                                    continue_from_epoch))  # restore previous graph to continue operations
 
     best_val_accuracy = 0.
-    with tqdm.tqdm(total=epochs) as epoch_pbar:
+    with tqdm.tqdm(total=epochs-start_epoch) as epoch_pbar:
         for e in range(start_epoch, epochs):
             total_c_loss = 0.
             total_accuracy = 0.
diff --git a/emnist_network_trainer.py b/emnist_network_trainer.py
index f33a436..1af1956 100644
--- a/emnist_network_trainer.py
+++ b/emnist_network_trainer.py
@@ -83,7 +83,7 @@ with tf.Session() as sess:
                                                    continue_from_epoch))  # restore previous graph to continue operations
 
     best_val_accuracy = 0.
-    with tqdm.tqdm(total=epochs) as epoch_pbar:
+    with tqdm.tqdm(total=epochs - start_epoch) as epoch_pbar:
         for e in range(start_epoch, epochs):
             total_c_loss = 0.
             total_accuracy = 0.
diff --git a/gpu_cluster_environment_script.sh b/gpu_cluster_environment_script.sh
new file mode 100644
index 0000000..5d9cd91
--- /dev/null
+++ b/gpu_cluster_environment_script.sh
@@ -0,0 +1,20 @@
+#!/bin/sh
+export CUDA_HOME=/opt/cuda-8.0.44
+
+export CUDNN_HOME=/opt/cuDNN-6.0_8.0
+
+export STUDENT_ID=sxxxxxx
+
+export LD_LIBRARY_PATH=${CUDNN_HOME}/lib64:${CUDA_HOME}/lib64:$LD_LIBRARY_PATH
+
+export LIBRARY_PATH=${CUDNN_HOME}/lib64:$LIBRARY_PATH
+
+export CPATH=${CUDNN_HOME}/include:$CPATH
+
+export PATH=${CUDA_HOME}/bin:${PATH}
+
+export PYTHON_PATH=$PATH
+
+# Activate the relevant virtual environment:
+
+source /home/${STUDENT_ID}/miniconda3/bin/activate mlp
diff --git a/msd10_network_trainer.py b/msd10_network_trainer.py
index 584347b..e379b4c 100644
--- a/msd10_network_trainer.py
+++ b/msd10_network_trainer.py
@@ -84,7 +84,7 @@ with tf.Session() as sess:
                                                    continue_from_epoch))  # restore previous graph to continue operations
 
     best_val_accuracy = 0.
-    with tqdm.tqdm(total=epochs) as epoch_pbar:
+    with tqdm.tqdm(total=epochs - start_epoch) as epoch_pbar:
         for e in range(start_epoch, epochs):
             total_c_loss = 0.
             total_accuracy = 0.
diff --git a/msd25_network_trainer.py b/msd25_network_trainer.py
index bc94997..ea86b8c 100644
--- a/msd25_network_trainer.py
+++ b/msd25_network_trainer.py
@@ -83,7 +83,7 @@ with tf.Session() as sess:
                                                    continue_from_epoch))  # restore previous graph to continue operations
 
     best_val_accuracy = 0.
-    with tqdm.tqdm(total=epochs) as epoch_pbar:
+    with tqdm.tqdm(total=epochs - start_epoch) as epoch_pbar:
         for e in range(start_epoch, epochs):
             total_c_loss = 0.
             total_accuracy = 0.