I don't obtain the model save_epoch_12

[KillerSH]

During training I get 2 warnings:

and

I don't know if this is the reason why I don't obtain the directory save_epoch_12. After training I don't obtain any save_epoch_12. In TrainSettings.py I set as you said PATH_TO_SAVE_MODEL and the only thing I obtain in that location are 3 directories: test, train and validation. Test is empty. Train contains:

Validation contains:

The following is my training log: ###############################################

WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0. For more information, please see:

• https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md
• https://github.com/tensorflow/addons If you depend on functionality not listed there, please file an issue.

Using Network: src.net.G2D19_P2OF_ResHB_1LSTM

2019-03-17 17:39:19.011373: W tensorflow/core/graph/graph_constructor.cc:1272] Importing a graph with a lower producer version 24 into an existing graph with producer version 27. Shape inference will have run different parts of the graph with different producer versions. WARNING:tensorflow:From C:\Users\stefa\Anaconda3\envs\DeepLearning\lib\site-packages\tensorflow\python\framework\op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version. Instructions for updating: Colocations handled automatically by placer. In OpticalFlow: pool2.shape = (?, 112, 112, 32) after transpose, out.shape = (?, 112, 112, 32, 2) before Conv2, out.shape = (?, 112, 112, 64) before Fc, out.shape = (?, 28, 28, 64) WARNING:tensorflow:From C:\Users\stefa\Downloads\ViolenceDetection-master\ViolenceDetection-master\src\net\G2D19_P2OF_ResHB_1LSTM.py:217: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version. Instructions for updating: Please use rate instead of keep_prob. Rate should be set to rate = 1 - keep_prob. Fc final.shape = (?, 1024) opticalFlowOut.shape = (?, 1024) In CNN: darknetOutput.shape = (?, 7, 7, 1024) after transpose, out.shape = (?, 7, 7, 1024, 2) before ConcatConv, out.shape = (?, 7, 7, 2048) before Fc, out.shape = (?, 7, 7, 128) Fc out.shape = (?, 1024) In Concat: before Fc, out.shape = (?, 2048) before LSTM, shape = (?, ?, 1024) WARNING:tensorflow:From C:\Users\stefa\Downloads\ViolenceDetection-master\ViolenceDetection-master\src\layers\RNN.py:14: LSTMCell.init (from tensorflow.python.ops.rnn_cell_impl) is deprecated and will be removed in a future version. Instructions for updating: This class is equivalent as tf.keras.layers.LSTMCell, and will be replaced by that in Tensorflow 2.0. WARNING:tensorflow:From C:\Users\stefa\Downloads\ViolenceDetection-master\ViolenceDetection-master\src\layers\RNN.py:27: dynamic_rnn (from tensorflow.python.ops.rnn) is deprecated and will be removed in a future version. Instructions for updating: Please use keras.layers.RNN(cell), which is equivalent to this API

WARNING:tensorflow:From C:\Users\stefa\Downloads\ViolenceDetection-master\ViolenceDetection-master\src\Classifier.py:34: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version. Instructions for updating:

Future major versions of TensorFlow will allow gradients to flow into the labels input on backprop by default.

See tf.nn.softmax_cross_entropy_with_logits_v2.

Reading Training set... WARNING:tensorflow:From C:\Users\stefa\Anaconda3\envs\DeepLearning\lib\site-packages\tensorflow\python\ops\math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version. Instructions for updating: Use tf.cast instead. Done.

Reading Validation set... Done.

Reading Test set... Done.

2019-03-17 17:40:15.224395: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX 2019-03-17 17:40:16.114667: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: name: GeForce GT 710M major: 2 minor: 1 memoryClockRate(GHz): 1.55 pciBusID: 0000:01:00.0 totalMemory: 1.00GiB freeMemory: 823.06MiB 2019-03-17 17:40:16.130714: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1483] Ignoring visible gpu device (device: 0, name: GeForce GT 710M, pci bus id: 0000:01:00.0, compute capability: 2.1) with Cuda compute capability 2.1. The minimum required Cuda capability is 3.0. 2019-03-17 17:40:16.147362: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix: 2019-03-17 17:40:16.155323: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990] 0 2019-03-17 17:40:16.162249: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0: N Path to save mode: C:\Users\stefa\Downloads\ViolenceDetection-master\ViolenceDetection-master\Model

Start Training...

2019-03-17 17:41:13.257704: W tensorflow/core/framework/allocator.cc:124] Allocation of 2055208960 exceeds 10% of system memory. 2019-03-17 17:41:36.646958: W tensorflow/core/framework/allocator.cc:124] Allocation of 2055208960 exceeds 10% of system memory. 2019-03-17 17:41:51.127135: W tensorflow/core/framework/allocator.cc:124] Allocation of 1027604480 exceeds 10% of system memory. 2019-03-17 17:42:58.710324: W tensorflow/core/framework/allocator.cc:124] Allocation of 1027604480 exceeds 10% of system memory. 2019-03-17 17:45:50.314111: W tensorflow/core/framework/allocator.cc:124] Allocation of 2055208960 exceeds 10% of system memory.

################################################## end

I used the hockey dataset for training and I have: CPU: I7 - 3537U dual core 2.2 GHz GPU: NVidia Geforce 710M , VRAM: 1 GB RAM: 8 GB

[JoshuaPiinRueyPan]

It seems that your Graphic Card does not have enough memory. I use NV GTX 1080 Ti, and while training it consumes nearly 10 GB memory...