I’ve figured it out myself: TL;DR: Make sure your loss magnitude is independent of your mini-batch size. The long explanation: In my case the issue was Keras-specific after all. Maybe the solution to this problem will be useful for someone at some point. It turns out that Keras divides the loss by the mini-batch size. … Read more
Yes that can be done. Its explained in the paper ‘Striving for simplicity: The all convolutional net’ https://arxiv.org/pdf/1412.6806.pdf. Quote from the paper: ‘We find that max-pooling can simply be replaced by a convolutional layer with increased stride without loss in accuracy on several image recognition benchmarks’
Unfortunately what you want to do is impossible with Keras … I’ve also struggle a lot of time on this problems and the only way is to dive into the rabbit hole and work with Tensorflow directly to do LSTM rolling prediction. First, to be clear on terminology, batch_size usually means number of sequences that … Read more
In my understanding, the “backbone” refers to the feature extracting network which is used within the DeepLab architecture. This feature extractor is used to encode the network’s input into a certain feature representation. The DeepLab framework “wraps” functionalities around this feature extractor. By doing so, the feature extractor can be exchanged and a model can … Read more
Since you trained your model on mini-batches, your input is a tensor of shape [batch_size, image_width, image_height, number_of_channels]. When predicting, you have to respect this shape even if you have only one image. Your input should be of shape: [1, image_width, image_height, number_of_channels]. You can do this in numpy easily. Let’s say you have a … Read more
Models and layers have special method for that purpose: model.count_params() Also, to get a short summary of each layer dimensions and parameters, you might find useful the following method model.summary()
The documentation is perhaps overly laconic, given that the terms used are fairly niche. In CUDA terms, pinned memory does not mean GPU memory but non-paged CPU memory. The benefits and rationale are provided here, but the gist of it is that this flag allows the x.cuda() operation (which you still have to execute as … Read more
The LSTM layer and the TimeDistributed wrapper are two different ways to get the “many to many” relationship that you want. LSTM will eat the words of your sentence one by one, you can chose via “return_sequence” to outuput something (the state) at each step (after each word processed) or only output something after the … Read more
The input shape you have defined is the shape of a single sample. The model itself expects some array of samples as input (even if its an array of length 1). Your output really should be 4-d, with the 1st dimension to enumerate the samples. i.e. for a single image you should return a shape … Read more
From model documentation: loss: String (name of objective function) or objective function. See losses. If the model has multiple outputs, you can use a different loss on each output by passing a dictionary or a list of losses. The loss value that will be minimized by the model will then be the sum of all … Read more