TFRecords - How to Read and Write Machine Learning Training Data

preview_player
Показать описание
How to use TFRecords to read and write training date for machine learning models. We're using TensorFlow 2.0 in this video.

TFRecords are a file format consisting of serialized data that's quick and efficient to read turning training time.

-----------------------------
-----------------------------

Рекомендации по теме
Комментарии
Автор

Thanks for the great video! could you help clarify what you mean by "they are serialized so we can read the data back out very quickly" (0:13) ? aren't other file formats like CSV or JSON serialized and deserialization is needed when reading the data back to memory?

joeykao
Автор

too good...Youtube should highlight your video

abhimalyachowdhury
Автор

Hi. it is good tutorial. how to handle images data? image data need two dimension array

cendradevayanaputra
Автор

The code could be a little bit bigger, but great content

Leibniz_
Автор

You actually created 10 examples, each containing a single feature whose value is a vector with 10 elements, not 10 features containing scalar values.

noli-timere-crede-tantum
Автор

what is the blue symbol ? is it an equal ?

geordi-gabrielrenauddumoul
Автор

No background music please - rest is awesome!

martinchristoph
join shbcf.ru