質問編集履歴
2
他の実例を加えました
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -113,3 +113,39 @@
|
|
|
113
113
|
（エラーメッセージ)
|
|
114
114
|
|
|
115
115
|
RuntimeError: $ Torch: not enough memory: you tried to allocate 0GB. Buy new RAM! at /opt/conda/conda-bld/pytorch_1512387374934/work/torch/lib/TH/THGeneral.c:246
|
|
116
|
+
|
|
117
|
+
|
|
118
|
+
|
|
119
|
+
```
|
|
120
|
+
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
ちなみになのですが、以下のコードは同じ環境で実行しても問題なく学習が進んでいます。
|
|
124
|
+
|
|
125
|
+
```python
|
|
126
|
+
|
|
127
|
+
for _, (images,img_masks) in enumerate(train_batch):
|
|
128
|
+
|
|
129
|
+
gen_optimizer.zero_grad()
|
|
130
|
+
|
|
131
|
+
|
|
132
|
+
|
|
133
|
+
x = Variable(images).cuda(0)
|
|
134
|
+
|
|
135
|
+
y_ = Variable(img_masks.mean(dim=1,keepdim=True)).cuda(0)
|
|
136
|
+
|
|
137
|
+
y = generator.forward(x).mean(dim=1,keepdim=True).cuda(0)
|
|
138
|
+
|
|
139
|
+
|
|
140
|
+
|
|
141
|
+
loss = recon_loss_func(y,y_) # nn.MSELoss()です。
|
|
142
|
+
|
|
143
|
+
file.write(str(loss.data[0])+"\n")
|
|
144
|
+
|
|
145
|
+
loss_record.append(loss.data[0])
|
|
146
|
+
|
|
147
|
+
loss.backward()
|
|
148
|
+
|
|
149
|
+
gen_optimizer.step()
|
|
150
|
+
|
|
151
|
+
```
|
1
出力を追記しました。
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -83,3 +83,33 @@
|
|
|
83
83
|
break
|
|
84
84
|
|
|
85
85
|
```
|
|
86
|
+
|
|
87
|
+
|
|
88
|
+
|
|
89
|
+
エラーメッセージを追記します。
|
|
90
|
+
|
|
91
|
+
出力
|
|
92
|
+
|
|
93
|
+
```
|
|
94
|
+
|
|
95
|
+
0
|
|
96
|
+
|
|
97
|
+
1
|
|
98
|
+
|
|
99
|
+
2
|
|
100
|
+
|
|
101
|
+
3
|
|
102
|
+
|
|
103
|
+
4
|
|
104
|
+
|
|
105
|
+
5
|
|
106
|
+
|
|
107
|
+
6
|
|
108
|
+
|
|
109
|
+
7
|
|
110
|
+
|
|
111
|
+
0 time loss: 0.8070976138114929
|
|
112
|
+
|
|
113
|
+
（エラーメッセージ)
|
|
114
|
+
|
|
115
|
+
RuntimeError: $ Torch: not enough memory: you tried to allocate 0GB. Buy new RAM! at /opt/conda/conda-bld/pytorch_1512387374934/work/torch/lib/TH/THGeneral.c:246
|