teratail
回答率
85.48%
質問するログイン新規登録
トップに関する質問Chainerで独自データセットを用いた学習が始まりません…。

編集履歴

質問編集履歴

6

cupy

2018/11/21 00:45

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -120,7 +120,7 @@
120
120
 
121
121
 
122
122
 
123
- data は 3次元のデータです。活性化関数を恒等関数にしてるのはわざとです。GPUは使いません。このコードを実行すると次のようなエラーがでるのですが、原因が全くわかりません。なぜでしょうか。
123
+ data は 3次元のデータです。活性化関数を恒等関数にしてるのはわざとです。このコードを実行すると次のようなエラーがでるのですが、原因が全くわかりません。なぜでしょうか。
124
124
 
125
125
 
126
126
 
@@ -250,7 +250,7 @@
250
250
 
251
251
 
252
252
 
253
- ちなみに、CUPTのエラーが出てるので、GPU関連かなと思い、
253
+ ちなみに cupyのエラーが出てるので、GPU関連かなと思い、
254
254
 
255
255
 
256
256
 

5

EDIT

2018/11/21 00:45

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -266,6 +266,114 @@
266
266
 
267
267
 
268
268
 
269
+ Exception in main training loop: unsupported operand type(s) for *: 'bool' and 'NoneType'
270
+
271
+ Traceback (most recent call last):
272
+
273
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
274
+
275
+ update()
276
+
277
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
278
+
279
+ self.update_core()
280
+
281
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
282
+
283
+ optimizer.update(loss_func, in_arrays)
284
+
285
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
286
+
287
+ loss.backward(loss_scale=self._loss_scale)
288
+
289
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
290
+
291
+ self._backward_main(retain_grad, loss_scale)
292
+
293
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
294
+
295
+ target_input_indexes, out_grad, in_grad)
296
+
297
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
298
+
299
+ gxs = self.backward(target_input_indexes, grad_outputs)
300
+
301
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
302
+
303
+ return ReLUGrad2(y).apply((gy,))
304
+
305
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
306
+
307
+ outputs = self.forward(in_data)
308
+
309
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
310
+
311
+ return self.forward_cpu(inputs)
312
+
313
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
314
+
315
+ y = (self.b > 0) * inputs[0]
316
+
317
+ Will finalize trainer extensions and updater before reraising the exception.
318
+
319
+ Traceback (most recent call last):
320
+
321
+ File "AC.py", line 70, in <module>
322
+
323
+ trainer.run()
324
+
325
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 322, in run
326
+
327
+ six.reraise(*sys.exc_info())
328
+
329
+ File "/home/****/.local/lib/python3.5/site-packages/six.py", line 693, in reraise
330
+
331
+ raise value
332
+
333
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
334
+
335
+ update()
336
+
337
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
338
+
339
+ self.update_core()
340
+
341
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
342
+
343
+ optimizer.update(loss_func, in_arrays)
344
+
345
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
346
+
347
+ loss.backward(loss_scale=self._loss_scale)
348
+
349
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
350
+
351
+ self._backward_main(retain_grad, loss_scale)
352
+
353
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
354
+
355
+ target_input_indexes, out_grad, in_grad)
356
+
357
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
358
+
359
+ gxs = self.backward(target_input_indexes, grad_outputs)
360
+
361
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
362
+
363
+ return ReLUGrad2(y).apply((gy,))
364
+
365
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
366
+
367
+ outputs = self.forward(in_data)
368
+
369
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
370
+
371
+ return self.forward_cpu(inputs)
372
+
373
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
374
+
375
+ y = (self.b > 0) * inputs[0]
376
+
269
377
  TypeError: unsupported operand type(s) for *: 'bool' and 'NoneType'
270
378
 
271
379
 

4

EDIT

2018/11/21 00:44

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -108,160 +108,160 @@
108
108
 
109
109
 
110
110
 
111
+ updater = training.StandardUpdater(train_iter,optimizer,device=0)
112
+
113
+ trainer = training.Trainer(updater,(n_epoch,"epoch"),out="result")
114
+
115
+
116
+
117
+ trainer.run()
118
+
119
+ ```
120
+
121
+
122
+
123
+ data は 3次元のデータです。活性化関数を恒等関数にしてるのはわざとです。GPUは使いません。このコードを実行すると次のようなエラーがでるのですが、原因が全くわかりません。なぜでしょうか。
124
+
125
+
126
+
127
+ Exception in main training loop: Unsupported type <class 'NoneType'>
128
+
129
+ Traceback (most recent call last):
130
+
131
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
132
+
133
+ update()
134
+
135
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
136
+
137
+ self.update_core()
138
+
139
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
140
+
141
+ optimizer.update(loss_func, in_arrays)
142
+
143
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
144
+
145
+ loss.backward(loss_scale=self._loss_scale)
146
+
147
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
148
+
149
+ self._backward_main(retain_grad, loss_scale)
150
+
151
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
152
+
153
+ target_input_indexes, out_grad, in_grad)
154
+
155
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
156
+
157
+ gxs = self.backward(target_input_indexes, grad_outputs)
158
+
159
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
160
+
161
+ return ReLUGrad2(y).apply((gy,))
162
+
163
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
164
+
165
+ outputs = self.forward(in_data)
166
+
167
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
168
+
169
+ return self.forward_cpu(inputs)
170
+
171
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
172
+
173
+ y = (self.b > 0) * inputs[0]
174
+
175
+ File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
176
+
177
+ File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
178
+
179
+ File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
180
+
181
+ Will finalize trainer extensions and updater before reraising the exception.
182
+
183
+ Traceback (most recent call last):
184
+
185
+ File "AC.py", line 70, in <module>
186
+
187
+ trainer.run()
188
+
189
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 322, in run
190
+
191
+ six.reraise(*sys.exc_info())
192
+
193
+ File "/home/****/.local/lib/python3.5/site-packages/six.py", line 693, in reraise
194
+
195
+ raise value
196
+
197
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
198
+
199
+ update()
200
+
201
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
202
+
203
+ self.update_core()
204
+
205
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
206
+
207
+ optimizer.update(loss_func, in_arrays)
208
+
209
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
210
+
211
+ loss.backward(loss_scale=self._loss_scale)
212
+
213
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
214
+
215
+ self._backward_main(retain_grad, loss_scale)
216
+
217
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
218
+
219
+ target_input_indexes, out_grad, in_grad)
220
+
221
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
222
+
223
+ gxs = self.backward(target_input_indexes, grad_outputs)
224
+
225
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
226
+
227
+ return ReLUGrad2(y).apply((gy,))
228
+
229
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
230
+
231
+ outputs = self.forward(in_data)
232
+
233
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
234
+
235
+ return self.forward_cpu(inputs)
236
+
237
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
238
+
239
+ y = (self.b > 0) * inputs[0]
240
+
241
+ File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
242
+
243
+ File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
244
+
245
+ File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
246
+
247
+ TypeError: Unsupported type <class 'NoneType'>
248
+
249
+
250
+
251
+
252
+
253
+ ちなみに、CUPTのエラーが出てるので、GPU関連かなと思い、
254
+
255
+
256
+
257
+ ```ここに言語を入力
258
+
111
259
  updater = training.StandardUpdater(train_iter,optimizer,device=-1)
112
260
 
113
- trainer = training.Trainer(updater,(n_epoch,"epoch"),out="result")
114
-
115
-
116
-
117
- trainer.run()
118
-
119
261
  ```
120
262
 
121
263
 
122
264
 
123
- data は 3次元のデータです。活性化関数を恒等関数にしてるのはわざとです。GPUは使いません。このコードを実行すると次のようなエラーがでるのですが、原因が全くわかりません。なぜでしょうか。
124
-
125
-
126
-
127
- Exception in main training loop: Unsupported type <class 'NoneType'>
128
-
129
- Traceback (most recent call last):
130
-
131
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
132
-
133
- update()
134
-
135
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
136
-
137
- self.update_core()
138
-
139
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
140
-
141
- optimizer.update(loss_func, in_arrays)
142
-
143
- File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
144
-
145
- loss.backward(loss_scale=self._loss_scale)
146
-
147
- File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
148
-
149
- self._backward_main(retain_grad, loss_scale)
150
-
151
- File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
152
-
153
- target_input_indexes, out_grad, in_grad)
154
-
155
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
156
-
157
- gxs = self.backward(target_input_indexes, grad_outputs)
158
-
159
- File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
160
-
161
- return ReLUGrad2(y).apply((gy,))
162
-
163
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
164
-
165
- outputs = self.forward(in_data)
166
-
167
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
168
-
169
- return self.forward_cpu(inputs)
170
-
171
- File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
172
-
173
- y = (self.b > 0) * inputs[0]
174
-
175
- File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
176
-
177
- File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
178
-
179
- File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
180
-
181
- Will finalize trainer extensions and updater before reraising the exception.
182
-
183
- Traceback (most recent call last):
184
-
185
- File "AC.py", line 70, in <module>
186
-
187
- trainer.run()
188
-
189
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 322, in run
190
-
191
- six.reraise(*sys.exc_info())
192
-
193
- File "/home/****/.local/lib/python3.5/site-packages/six.py", line 693, in reraise
194
-
195
- raise value
196
-
197
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
198
-
199
- update()
200
-
201
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
202
-
203
- self.update_core()
204
-
205
- File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
206
-
207
- optimizer.update(loss_func, in_arrays)
208
-
209
- File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
210
-
211
- loss.backward(loss_scale=self._loss_scale)
212
-
213
- File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
214
-
215
- self._backward_main(retain_grad, loss_scale)
216
-
217
- File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
218
-
219
- target_input_indexes, out_grad, in_grad)
220
-
221
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
222
-
223
- gxs = self.backward(target_input_indexes, grad_outputs)
224
-
225
- File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
226
-
227
- return ReLUGrad2(y).apply((gy,))
228
-
229
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
230
-
231
- outputs = self.forward(in_data)
232
-
233
- File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
234
-
235
- return self.forward_cpu(inputs)
236
-
237
- File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
238
-
239
- y = (self.b > 0) * inputs[0]
240
-
241
- File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
242
-
243
- File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
244
-
245
- File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
246
-
247
- TypeError: Unsupported type <class 'NoneType'>
248
-
249
-
250
-
251
-
252
-
253
- ちなみに、CUPTのエラーが出てるので、GPU関連かなと思い、
254
-
255
-
256
-
257
- ```ここに言語を入力
258
-
259
- updater = training.StandardUpdater(train_iter,optimizer,device=-1)
260
-
261
- ```
262
-
263
-
264
-
265
265
  としてみたのですが、今度は
266
266
 
267
267
 

3

追記

2018/11/21 00:41

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -245,3 +245,29 @@
245
245
  File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
246
246
 
247
247
  TypeError: Unsupported type <class 'NoneType'>
248
+
249
+
250
+
251
+
252
+
253
+ ちなみに、CUPTのエラーが出てるので、GPU関連かなと思い、
254
+
255
+
256
+
257
+ ```ここに言語を入力
258
+
259
+ updater = training.StandardUpdater(train_iter,optimizer,device=-1)
260
+
261
+ ```
262
+
263
+
264
+
265
+ としてみたのですが、今度は
266
+
267
+
268
+
269
+ TypeError: unsupported operand type(s) for *: 'bool' and 'NoneType'
270
+
271
+
272
+
273
+ というエラーがでます。

2

x を h2に直した。

2018/11/21 00:18

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -24,7 +24,7 @@
24
24
 
25
25
  h1 = self.l1(x)
26
26
 
27
- h2 = self.l2(x)
27
+ h2 = self.l2(h1)
28
28
 
29
29
 
30
30
 
@@ -124,142 +124,124 @@
124
124
 
125
125
 
126
126
 
127
-
128
-
129
- .local/lib/python3.5/site-packages/chainer/training/extensions/evaluator.py:107: UserWarning: The `repeat` property of the iterator {}
130
-
131
- warnings.warn(msg)
132
-
133
- Exception in main training loop:
127
+ Exception in main training loop: Unsupported type <class 'NoneType'>
134
-
135
- Invalid operation is performed in: LinearFunction (Forward)
136
-
137
-
138
-
139
- Expect: in_types[0].shape[1] == in_types[1].shape[1]
140
-
141
- Actual: 3 != 2
142
128
 
143
129
  Traceback (most recent call last):
144
130
 
145
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
131
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
146
132
 
147
133
  update()
148
134
 
149
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
135
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
150
136
 
151
137
  self.update_core()
152
138
 
153
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
139
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
154
140
 
155
141
  optimizer.update(loss_func, in_arrays)
156
142
 
157
- File "/home/***/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 650, in update
143
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
158
-
159
- loss = lossfun(*args, **kwds)
144
+
160
-
161
- File "AC.py", line 23, in __call__
145
+ loss.backward(loss_scale=self._loss_scale)
162
-
163
- h2 = F.relu(self.l2(x))
146
+
164
-
165
- File "/home/***/.local/lib/python3.5/site-packages/chainer/links/connection/linear.py", line 134, in __call__
147
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
148
+
166
-
149
+ self._backward_main(retain_grad, loss_scale)
150
+
151
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
152
+
167
- return linear.linear(x, self.W, self.b)
153
+ target_input_indexes, out_grad, in_grad)
154
+
168
-
155
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
156
+
157
+ gxs = self.backward(target_input_indexes, grad_outputs)
158
+
169
- File "/home/***/.local/lib/python3.5/site-packages/chainer/functions/connection/linear.py", line 234, in linear
159
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
170
-
160
+
171
- y, = LinearFunction().apply(args)
161
+ return ReLUGrad2(y).apply((gy,))
172
-
162
+
173
- File "/home/***/.local/lib/python3.5/site-packages/chainer/function_node.py", line 243, in apply
163
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
174
-
164
+
175
- self._check_data_type_forward(in_data)
165
+ outputs = self.forward(in_data)
176
-
166
+
177
- File "/home/***/.local/lib/python3.5/site-packages/chainer/function_node.py", line 328, in _check_data_type_forward
167
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
178
-
168
+
179
- self.check_type_forward(in_type)
169
+ return self.forward_cpu(inputs)
180
-
170
+
181
- File "/home/***/.local/lib/python3.5/site-packages/chainer/functions/connection/linear.py", line 23, in check_type_forward
171
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
182
-
172
+
183
- x_type.shape[1] == w_type.shape[1],
173
+ y = (self.b > 0) * inputs[0]
174
+
184
-
175
+ File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
176
+
177
+ File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
178
+
185
- File "/home/***/.local/lib/python3.5/site-packages/chainer/utils/type_check.py", line 524, in expect
179
+ File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
186
-
187
- expr.expect()
188
-
189
- File "/home/***/.local/lib/python3.5/site-packages/chainer/utils/type_check.py", line 482, in expect
190
-
191
- '{0} {1} {2}'.format(left, self.inv, right))
192
180
 
193
181
  Will finalize trainer extensions and updater before reraising the exception.
194
182
 
195
183
  Traceback (most recent call last):
196
184
 
197
- File "AC.py", line 72, in <module>
185
+ File "AC.py", line 70, in <module>
198
186
 
199
187
  trainer.run()
200
188
 
201
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 322, in run
189
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 322, in run
202
190
 
203
191
  six.reraise(*sys.exc_info())
204
192
 
205
- File "/home/***/.local/lib/python3.5/site-packages/six.py", line 693, in reraise
193
+ File "/home/****/.local/lib/python3.5/site-packages/six.py", line 693, in reraise
206
194
 
207
195
  raise value
208
196
 
209
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
197
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/trainer.py", line 308, in run
210
198
 
211
199
  update()
212
200
 
213
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
201
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 149, in update
214
202
 
215
203
  self.update_core()
216
204
 
217
- File "/home/***/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
205
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/training/updaters/standard_updater.py", line 164, in update_core
218
206
 
219
207
  optimizer.update(loss_func, in_arrays)
220
208
 
221
- File "/home/***/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 650, in update
209
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/optimizer.py", line 655, in update
222
-
223
- loss = lossfun(*args, **kwds)
210
+
224
-
225
- File "AC.py", line 23, in __call__
211
+ loss.backward(loss_scale=self._loss_scale)
226
-
227
- h2 = F.relu(self.l2(x))
212
+
228
-
229
- File "/home/***/.local/lib/python3.5/site-packages/chainer/links/connection/linear.py", line 134, in __call__
213
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 966, in backward
214
+
230
-
215
+ self._backward_main(retain_grad, loss_scale)
216
+
217
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/variable.py", line 1095, in _backward_main
218
+
231
- return linear.linear(x, self.W, self.b)
219
+ target_input_indexes, out_grad, in_grad)
220
+
232
-
221
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 548, in backward_accumulate
222
+
223
+ gxs = self.backward(target_input_indexes, grad_outputs)
224
+
233
- File "/home/***/.local/lib/python3.5/site-packages/chainer/functions/connection/linear.py", line 234, in linear
225
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 73, in backward
234
-
226
+
235
- y, = LinearFunction().apply(args)
227
+ return ReLUGrad2(y).apply((gy,))
236
-
228
+
237
- File "/home/***/.local/lib/python3.5/site-packages/chainer/function_node.py", line 243, in apply
229
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 258, in apply
238
-
230
+
239
- self._check_data_type_forward(in_data)
231
+ outputs = self.forward(in_data)
240
-
232
+
241
- File "/home/***/.local/lib/python3.5/site-packages/chainer/function_node.py", line 328, in _check_data_type_forward
233
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/function_node.py", line 368, in forward
242
-
234
+
243
- self.check_type_forward(in_type)
235
+ return self.forward_cpu(inputs)
244
-
236
+
245
- File "/home/***/.local/lib/python3.5/site-packages/chainer/functions/connection/linear.py", line 23, in check_type_forward
237
+ File "/home/****/.local/lib/python3.5/site-packages/chainer/functions/activation/relu.py", line 97, in forward_cpu
246
-
238
+
247
- x_type.shape[1] == w_type.shape[1],
239
+ y = (self.b > 0) * inputs[0]
240
+
248
-
241
+ File "cupy/core/core.pyx", line 1310, in cupy.core.core.ndarray.__mul__
242
+
243
+ File "cupy/core/elementwise.pxi", line 753, in cupy.core.core.ufunc.__call__
244
+
249
- File "/home/***/.local/lib/python3.5/site-packages/chainer/utils/type_check.py", line 524, in expect
245
+ File "cupy/core/elementwise.pxi", line 68, in cupy.core.core._preprocess_args
250
-
251
- expr.expect()
246
+
252
-
253
- File "/home/***/.local/lib/python3.5/site-packages/chainer/utils/type_check.py", line 482, in expect
254
-
255
- '{0} {1} {2}'.format(left, self.inv, right))
256
-
257
- chainer.utils.type_check.InvalidType:
258
-
259
- Invalid operation is performed in: LinearFunction (Forward)
260
-
261
-
262
-
263
- Expect: in_types[0].shape[1] == in_types[1].shape[1]
247
+ TypeError: Unsupported type <class 'NoneType'>
264
-
265
- Actual: 3 != 2

1

GPU

2018/11/20 23:23

投稿

physics303
physics303

スコア89

test CHANGED
File without changes
test CHANGED
@@ -108,14 +108,10 @@
108
108
 
109
109
 
110
110
 
111
- updater = training.StandardUpdater(train_iter,optimizer,device=0)
111
+ updater = training.StandardUpdater(train_iter,optimizer,device=-1)
112
112
 
113
113
  trainer = training.Trainer(updater,(n_epoch,"epoch"),out="result")
114
114
 
115
- trainer.extend(extensions.Evaluator(train_iter,model,device=0),name="val")
116
-
117
-
118
-
119
115
 
120
116
 
121
117
  trainer.run()