Bartosz Ptak

My interesting cases


24 Jul 2019

Keras: how to load own loss and metric

ValueError: Unknown loss function:loss - I received this error by trying to load the model. It happened as I created...

23 Jul 2019

Keras: own loss and metric in the model

Today a very simple thing: create your own loss and metric functions in Keras (Tensorflow backend). The most important thing...

17 Jul 2019

Tensorflow: save, load and use protobuf model

Tensorflow enables serialization and deserialization of models in many ways. This article shows how to use the Protocol Buffer to...

02 Jul 2019

Multiprocessing: using the queue to parallel calculations

The second article on the parallelism of calculations describes the use of the queue. Data is delivered in batches and...

14 Jun 2019

Gensim: convert GloVe to word2vec model

GloVe means Global Vectors for Word Representation. The authors provide pre-trained word vectors models learned on such collections as: Wikipedia...

04 Jun 2019

Tenforflow: convert keras model to Tensorflow Lite

In today’s article, I will briefly show you how to convert the Keras model (.h5) to Tensorflow-Lite (.tflite). This action...

29 May 2019

Horovod: distributed training with Tensorflow from Uber

Horovod is a great tool for distributed training. This Uber-made framework is built on the basis of MPI and TCP...

21 May 2019

Multiprocessing: using the pool to parallel calculations

Modern processors have many cores and threads. Why not use them for calculations in Python? Many frameworks have built-in support...