mse

I've been trying to translate some PyTorch code to TensorFlow 2, but the TF2 code is around 10 times slower. I've tried looking at where this might come from, and as far as I can tell it comes from the tape.gradient call (performance was the same with keras' .fit function). I've tried to use different data loaders, ways of declaring the model, installations, etc... and the results have been consistent.

Any explanation / solution as to why this is happening would be much appreciated.

Here is a minimalist version of the TF2 code:

Background

I'm currently trying to implement a DDPG framework to control a simple car agent. At first, the car agent would only need to learn how to reach the end of a straight path as quickly as possible by adjusting its acceleration. This task was simple enough, so I decided to introduce an additional steering action as well. I updated my observation and action spaces accordingly.

The lines below are the for loop that runs each episode:

I have a CNN-LSTM that looks as follows;

this is my Define validate function
when I load the model and start prediction using this code I have received the error using PyTorch.and after this, I am iterating through the epoch loop and batch loop and I landed with this error.

I'm implementing SRGAN (and am not very experienced in this field), which uses a pre-trained VGG19 model to extract features. The following code was working fine on Keras 2.1.2 and tf 1.15.0 till yesterday. then it started throwing an "AttributeError: module 'keras.utils.generic_utils' has no attribute 'populate_dict_with_module_objects'" So i updated the keras version to 2.4.3 and tf to 2.5.0. but then its showing a "Input 0 of layer fc1 is incompatible with the layer: expected axis -1 of input shape to have value 25088 but received input with shape (None, 32768)" on the following line

I have a concern in understanding the Cartpole code as an example for Deep Q Learning. The DQL Agent part of the code as follow:

I have an object with a structure similar to:

I am trying to implement a Keras regression model that learns to set some parameters, e.g there are some parameters in input and a set of unrelated outputs, coherent with the inputs (e.g. similar inputs give similar outputs in the training set, and there is partial linearity between some inputs and some outputs). Inputs and outputs are normalized, since the parameters have different units.

The training phase results in a mse of ~ 0.48 and the predictions are rather sensible.

This is the model:

I tried to refer to below links to use data.table package to calculate mape,smape,mse,rmse, its working fine with below codes when run in chunk but become error when knit2html. Somebody shade me a light?

• https://tysonbarrett.com/jekyll/update/2019/10/06/datatable_memory
• http://brooksandrew.github.io/simpleblog/articles/advanced-data-table