Yes is it possible, at least with version 2.0.4 (don’t know about earlier version). The instance of ImageDataGenerator().flow_from_directory(…) has an attribute with filenames which is a list of all the files in the order the generator yields them and also an attribute batch_index. So you can do it like this: datagen = ImageDataGenerator() gen = … Read more
The correct way to use the advanced activations like PReLU is to use it with add() method and not wrapping it using Activation class. Example: model = Sequential() act = keras.layers.advanced_activations.PReLU(init=”zero”, weights=None) model.add(Dense(64, input_dim=14, init=”uniform”)) model.add(act)
Try the following callback: class TimeHistory(keras.callbacks.Callback): def on_train_begin(self, logs={}): self.times = [] def on_epoch_begin(self, batch, logs={}): self.epoch_time_start = time.time() def on_epoch_end(self, batch, logs={}): self.times.append(time.time() – self.epoch_time_start) Then: time_callback = TimeHistory() model.fit(…, callbacks=[…, time_callback],…) times = time_callback.times In this case times should store the epoch computation times.
Part 1: The problem is indeed the datatype of your input. By default your keras model expects float32 but you are passing a float64. You can either change the dtype of the model or change the input to float32. To change your model: def make_model(): net = tf.keras.Sequential() net.add(tf.keras.layers.Dense(4, activation=’relu’, dtype=”float32″)) net.add(tf.keras.layers.Dense(4, activation=’relu’)) net.add(tf.keras.layers.Dense(1)) return … Read more
The Flatten() operator unrolls the values beginning at the last dimension (at least for Theano, which is “channels first”, not “channels last” like TF. I can’t run TensorFlow in my environment). This is equivalent to numpy.reshape with ‘C’ ordering: ‘C’ means to read / write the elements using C-like index order, with the last axis … Read more
If your first 9 layers are consistently named between your original trained model and the new model, then you can use model.load_weights() with by_name=True. This will update weights only in the layers of your new model that have an identically named layer found in the original trained model. The name of the layer can be … Read more
I came up with your case. This is how I sort it out Install Anaconda Create a virtual environment – conda create -n tensorflow Go inside your virtual environment – (on macOS/Linux:) source activate tensorflow (on Windows: activate tensorflow) Inside that install tensorflow. You can install it using pip Finish install So then the next … Read more
Keras and TF themselves don’t use whole cores and capacity of CPU! If you are interested in using all 100% of your CPU then the multiprocessing.Pool basically creates a pool of jobs that need doing. The processes will pick up these jobs and run them. When a job is finished, the process will pick up … Read more
The problem here is unintentional broadcasting in the PyTorch training loop. The result of a nn.Linear operation always has shape [B,D], where B is the batch size and D is the output dimension. Therefore, in your mean_squared_error function ypred has shape [32,1] and ytrue has shape [32]. By the broadcasting rules used by NumPy and … Read more
From the recent Deep Learning book by Goodfellow et al., chapter 8: Minibatch sizes are generally driven by the following factors: Larger batches provide a more accurate estimate of the gradient, but with less than linear returns. Multicore architectures are usually underutilized by extremely small batches. This motivates using some absolute minimum batch size, below … Read more