I have a NN implemented in tensorflow that outputs a 1D tensor containing N float values in the interval (0,M). Would it be possible to transform it, using differentiable tensorflow operations, into a sparse 2D tensor of shape NxM, with one non-null value (e.g. 1) per row. The position of the non-null value of row ..

#### Category : tensorflow2.0

I am Tensorflow newbie. I have model generated using convNetKerasLarge.py and saved as tflite model. I am trying to test this saved model as follows import tensorflow as tf import numpy as np import glob from skimage.transform import resize from skimage import io # out of previously used training and test set start = 4001 ..

I am working on a CRNN model where the input is of sequence length 512. This means, in a single sample there are 512 time-frames. I would like to calculate the validation accuracy (categorical/binary accuracy) for each frame in a sample. Is this doable? If yes, then how can I do this? In standard implementations, ..

I am trying to use GPU with Tensorflow. My Tensorflow version is 2.4.1 and I am using Cuda version 11.2. Here is the output of nvidia-smi. +—————————————————————————–+ | NVIDIA-SMI 460.39 Driver Version: 460.39 CUDA Version: 11.2 | |——————————-+———————-+———————-+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage ..

I have been trying to train a model for detecting cats in photos using a CNN. And I have been using these datasets : https://www.kaggle.com/shaunthesheep/microsoft-catsvsdogs-dataset https://www.kaggle.com/balraj98/facades-dataset I have been getting this error: tensorflow.python.framework.errors_impl.InvalidArgumentError: Input size should match (header_size + row_size * abs_height) but they differ by 2 [[{{node decode_image/DecodeImage}}]] [[IteratorGetNext]] [Op:__inference_train_function_962] Function call stack: train_function ..

I have built a DNN for classifying the Fashion MNIST images using Keras (code below), but without the data pipeline.I want to incorporate tf.data pipeline in the DNN model for classifying the Fashion MNIST data.So how do I incorporate tf.data pipeline in the below DNN model ? Also how can I visualise the weights initialised ..

I need to generate images with a Sum Product Network (like a GAN) with Tensorflow 2 (and Python3.x). I found some libs like (https://github.com/pronobis/libspn-keras) or (https://github.com/SPFlow/SPFlow) But the last one, the examples are not for TF 2 sadly. My question is now, do they have a sample/generate function? I didnt found one. Source: Python-3x..

I read in an installation procedure that for 2.4 version of tensorflow , after installing the nvidia cuda toolkit and cuDNN it is enough to do "$ pip install tensorflow" (to access the GPU) and these days it is not necessary for doing "$ pip install tensorflow-gpu". Am I right about this , also by ..

I am training a CNN AlexNet model on a quite large dataset (DeepFashion) containing around 300,000 images. I resize the images to 96x96x3 and I am using GPU nvidia tesla K80 and 4 vCPUs, 15 GB memory on GCP (google cloud). The thing is that when I start the training it runs really well for ..

I’m a pretty new of Tensorflow. I created a sub-classed custom layer which has Variable. But I’m not sure how to visualize the Variable updating and the layer’s input and output on Tensorboard. Questions Basically, what are the best ways to get Variables in custom layers and the input/output of layers during the training process? ..

## Recent Comments