This message, related to the development of the theme, only displays on the localhost homepage to notify you of any important theme changes.

Version 2.0.0 - July 20, 2020

Below are the following changes that could be breaking changes for your site. For more details on any change, please refer to PR #154.

The major breaking change is:

  1. Users that have front matter that utilize images (backwards compatibility for featured and associated parameters still remains) will need to adjust from [images]="SRC" to the new format.
[[images]]
    src = "" // Link to image
    alt = "" // Alt text for image
    stretch = // Optional: See screenshots for referenced values and outcomes

If you utilize any of the following, there might be a breaking:

  1. User custom templates may require adjustment.
  2. User custom i18n languages, or custom templates referencing i18n translations may require adjustment.
  3. User custom template for comments will require adjustment if it uses the theme’s CSS and/or JS.
  4. User custom CSS may need to adjust due to a variety of class name changes and specificity changes.

While I realize this is inconvenient, I hope that it is worth it to you in the long run. Thanks for using the theme, and feel free to submit issues as needed.

Model Fan

在 Ubuntu 20.04 中安裝 CUDA 和 cuDNN.tw

1-Minute Read

硬體設備有了圖形處理單元（graphical processing units，GPU），當然也要有相對應的軟體。CUDA 就是 NVIDIA 發明用於 GPU 的平行運算平台和程式設計模型，有了 CUDA 我們才能在寫程式時享受到 GPU 帶來的高速運算能力。而 cuDNN.tw 是為了深層神經網路設計的 GPU 加速原式函式庫，常見的應用有：convolution、pooling 和 softmax 等。在前一個環境(亦是 Ubuntu)中我並沒有安裝，但我想反正這些也是常用之功能，NVIDIA 都大發慈悲的給予加速器，不用白不用，於是一起安裝。

在 Windows 中安裝 Ubuntu 20.04 雙系統

1-Minute Read

利用高階 NVIDIA 顯示卡做深度學習可以大幅提身運算效能，然而顯示卡在 windows 系統下會限制其計算能力。因此，這篇文章記錄在 windows 環境中安裝 Ubuntu 雙系統，並更新驅動程式。安裝其他深度學習所需程式則會在下一篇文章中介紹。

Recent Posts

Category

About

Model Fan is created by Tina, a Statistics PhD student focusing on spatial statistics and machine learning.