Siamese-tf2
Siamese-tf2 copied to clipboard
bubbliiiing
使用自己準備的資料集跑完一個epoch後就沒反應
您好: 請問我使用自己準備的梅爾頻譜圖資料集(圖片大小432x288)在訓練完一個epoch後就會沒有反應停在原地，可能是發生了甚麼狀況呢?
Number of devices: 1
Configurations:
----------------------------------------------------------------------
| keys | values|
----------------------------------------------------------------------
| model_path | model_data/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5|
| input_shape | [105, 105]|
| Init_Epoch | 0|
| Epoch | 100|
| batch_size | 64|
| Init_lr | 0.01|
| Min_lr | 0.0001|
| optimizer_type | sgd|
| momentum | 0.9|
| lr_decay_type | cos|
| save_period | 10|
| save_dir | logs|
| num_workers | 1|
| num_train | 900|
| num_val | 100|
----------------------------------------------------------------------
[Warning] 使用sgd优化器时，建议将训练总步长设置到30000以上。
[Warning] 本次运行的总训练数据量为900，batch_size为64，共训练100个Epoch，计算出总训练步长为1400。
[Warning] 由于总训练步长为1400，小于建议总步长30000，建议设置总世代为2143。
C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\optimizers\optimizer_v2\adam.py:110: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
super(Adam, self).__init__(name, **kwargs)
C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\optimizers\optimizer_v2\gradient_descent.py:108: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.
super(SGD, self).__init__(name, **kwargs)
Train on 900 samples, val on 100 samples, with batch size 64.
Epoch 1: LearningRateScheduler setting learning rate to 0.001.
Epoch 1/100
6/14 [===========>..................] - ETA: 10s - loss: 0.7094 - binary_accuracy: 0.4935WARNING:tensorflow:Callback method `on_train_batch_end` is slow compared to the batch time (batch time: 0.4824s vs `on_train_batch_end` time: 0.6985s). Check your callbacks.
14/14 [==============================] - ETA: 0s - loss: 0.7034 - binary_accuracy: 0.4933
手動Ctrl+c終止後的訊息:
Traceback (most recent call last):
File "c:\Users\little7\Downloads\Siamese-tf2-master\train.py", line 283, in <module>
model.fit(
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\utils\traceback_utils.py", line 64, in error_handler
return fn(*args, **kwargs)
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\training.py", line 1432, in fit
self._eval_data_handler = data_adapter.get_data_handler(
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\data_adapter.py", line 1401, in get_data_handler
return DataHandler(*args, **kwargs)
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\data_adapter.py", line 1151, in __init__
self._adapter = adapter_cls(
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\data_adapter.py", line 926, in __init__
super(KerasSequenceAdapter, self).__init__(
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\data_adapter.py", line 800, in __init__
peek, x = self._peek_and_restore(x)
File "C:\Users\little7\anaconda3\envs\tf2-gpu\lib\site-packages\keras\engine\data_adapter.py", line 937, in _peek_and_restore
return x[0], x
File "c:\Users\little7\Downloads\Siamese-tf2-master\utils\dataloader.py", line 44, in __getitem__
selected_path = self.train_lines[self.train_labels[:] == c]
KeyboardInterrupt
