So guys, I am using tensorflow==1.15.0 and python==3.7.10. I can’t upgrade to tensorflow==2.X because I am working upon the code which has been written in 1.X version. So, here is the complete Error:- Type of the model_obj: <class ‘CustomAutoencoder_tf21s.AutoEncoderModel’> Type of test_norm: <class ‘numpy.ndarray’> The test_norm is [[-0.8249858 -1.89925946 -0.19963089 -0.72175069 -0.08146654 0.17114502 -0.78429501 -0.41983836 ..
I have a task where have to quickly run N parallel simple optimizations with Adam optimizers. I have been doing this with tensorflow 1.x for a while but trying to update everything to 2.x or alternatively modern pytorch has resulted in much slower behavior. I constructed minimal(at least to best of my ability) examples demonstrating ..
how we can swap the tensor dimension e.g. if we have a tensor(1,64,14,20) and I want to swap the dimension axis to the tensor(1,20,14,64). I want to do it in TensorFlow without using transpose and Keras. Source: Python-3x..
I am using Tensorflow’s object detection api to train ssd model. I am using for that Tensorflow 1.15 in google colab and google drive for storage. As the training went the oldest checkpoints gets deleted but my problem is that google drive has deleted .data file of the latest checkpoint and I only have .meta ..
I am trying to train a model using tensorflow 1.15 with eager execution enabled. For train loss I am using train loss = mse_loss*args.lmbda + bits_per_pixel_loss I’ve defined the optimizer as below main_optimiser = tf.train.AdamOptimiser(learning_rate=1e-3) Upto here the code is working fine The error is coming in the minimizer part of the optimisation main_step = ..
Goal: I need to install Tensorflow 1 in order to use Object Detection. I have tried to look online for an answer I couldn’t find. I have python version 3.8.5 64 bit. I have also tried pip install tensorflow==1.15. As shown in the Docs: https://www.tensorflow.org/install/pip? Source: Python..
I found a pre trained model online saved using tensorflow v1 Saver object in here. My goal is to load this model and fine tune it, but I have never used tensorflow v1 before and I was wandering is it is possible to load this model and then fine tune it using tensorflow 2 or, ..
I have a use case of moving tf models to py3. Originally the model was trained with py 2.7 using TensorFlow 1.12. I read the same version of TF is supported in py 3.6. So if I directly load a model in py36, will it give the same prediction scores wrt reproducibility of the model. ..
I’m trying to get the same results using tf 1.15 and tf 2.4.1. Here are 2 examples that I’m expecting to produce similar results. The first example using tf.layers.conv2d() and the second using tf.keras.layers.Conv2D(). I tried using kernel_initializer=some_tf_initializer(seed=seed) and I also tried without using a kernel initializer, still the same. ========================================================================== TF1 import tensorflow as ..
I need to do this very same calculation in TF2 which ends up by calculating v1 and v2 which are softmax output and the same softmax output but calculated using a tf.train.ExponentialMovingAverage average() method. The original example can be found here at line 88 but I simplified the code in the example here for readability. ..