From c8bc2db12cddbf90b851798cd101632e7e9511ba Mon Sep 17 00:00:00 2001
From: Francois Fleuret <francois@fleuret.org>
Date: Fri, 16 Jun 2017 00:27:31 +0200
Subject: [PATCH] Make the name of the saved model more explicit.

---
 cnn-svrt.py | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/cnn-svrt.py b/cnn-svrt.py
index 084606a..a2ab1a3 100755
--- a/cnn-svrt.py
+++ b/cnn-svrt.py
@@ -107,6 +107,7 @@ class AfrozeShallowNet(nn.Module):
         self.conv3 = nn.Conv2d(16, 120, kernel_size=18)
         self.fc1 = nn.Linear(120, 84)
         self.fc2 = nn.Linear(84, 2)
+        self.name = 'shallownet'
 
     def forward(self, x):
         x = fn.relu(fn.max_pool2d(self.conv1(x), kernel_size=2))
@@ -117,6 +118,8 @@ class AfrozeShallowNet(nn.Module):
         x = self.fc2(x)
         return x
 
+######################################################################
+
 def train_model(model, train_set):
     batch_size = args.batch_size
     criterion = nn.CrossEntropyLoss()
@@ -178,7 +181,7 @@ for problem_number in range(1, 24):
         nb_parameters += p.numel()
     log_string('nb_parameters {:d}'.format(nb_parameters))
 
-    model_filename = 'model_' + str(problem_number) + '.param'
+    model_filename = model.name + '_' + str(problem_number) + '_' + str(train_set.nb_batches) + '.param'
 
     try:
         model.load_state_dict(torch.load(model_filename))
-- 
2.20.1