Placeholder Image

字幕列表 影片播放

  • SPEAKER: Comma separated values, or CSV,

  • is a commonly-used data format.

  • In this video, you'll see how to use CSV and Keras with Eager

  • Execution turned on so you can load data

  • for training a neural network and view it imperatively.

  • First of all, here's the URL where the data is stored.

  • Keras has a Get File utility, which

  • can download a file from a URL, so this

  • uses that to get the data and store it locally.

  • There are five columns in the data-- four

  • for features, one for the label.

  • So this code simply creates lists of each of these.

  • You'll need them later.

  • We need to specify how large the batches of data that we'll load

  • are.

  • This is how many records it will process at a time.

  • And the Make CSV Data Set utility

  • is where the magic happens.

  • You simply tell it the data, batch size, column, and labels,

  • and the number of epochs that you wanted to run for,

  • and it will load and slice the data for you.

  • We can inspect the data in the debugger

  • to see that at least something was loaded.

  • But if we want to see more, we can just

  • look at the next iter of the training data

  • set that was created, then print out the features or labels.

  • And as you can see in the output,

  • we can see that the features are loaded nicely

  • into lists for us.

  • So that's it for loading CSV into TensorFlow

  • using Keras Utilities.

  • To learn more about TensorFlow, visit tensorflow.org.

  • For more videos about TensorFlow,

  • click the Subscribe button.

  • And if you've any questions about this video,

  • please leave them in the comments below.

  • Thank you.

SPEAKER: Comma separated values, or CSV,

字幕與單字

單字即點即查 點擊單字可以查詢單字解釋

B1 中級

為Keras快速獲取CSV到數據集中(TensorFlow本週小技巧)。 (Quickly get CSV into datasets for Keras (TensorFlow Tip of the Week))

  • 3 0
    林宜悉 發佈於 2021 年 01 月 14 日
影片單字