Markdown Syntax
Markdown is a simple way to create rich documents using intuitive formatting hints. Unlike HTML, markdown is designed to be easy to read even in its raw form. When markdown is "rendered" in the web interface, it will use those hints to create an attractive visual representation. The description field of an Asset uses markdown to allow sharing of complex information.
The Basics
Typical markdown documents are paragraphs of text. When this text is shown to a user the text will automatically be reformatted to match the user's screen.
Text in a single block will be "re-flowed". Text with empty lines in between is considered a new paragraph, but multiple single line are considered to be part of the same block. Here are some examples:
Here is some sample text which will reflow. Include two extra spaces at the end to prevent text reflow. Here is another paragraph in this example which writes everything on the same very long line. This line will also reflow to look good on the user's screen.
This text is reflowed to render as such:
Here is some sample text which will reflow. Include two extra spaces at the end to prevent text reflow. Here is another paragraph in this example which writes everything on the same very long line. This line will also reflow to look good on the user's screen.
Headers (aka Titles and Section Titles)
The # character is used to create titles and subtitles for sections of text. You can think of additional # characters as the word "sub". So ### creates a sub-sub-title.
# Basic title Followed by text ## Sub title Followed by text ### Sub Sub title Followed by text #### 4th sub title Followed by text ##### 5th sub title -- slightly different way to call attention Followed by text
This renders as:
Basic title Followed by text Sub title Followed by text
Sub Sub title Followed by text 4th sub title Followed by text 5th sub title -- slightly different way to call attention Followed by text
Basic text formatting
Markdown uses simple characters to create textual effects.
You can apply **bold** (or **bold**) and _italics_ (or *italics*). You can also `quote` text.
This renders as:
You can apply bold (or bold) and italics (or italics). You can also quote text.
Lists
Ordered (numbered) and unordered (bullet) lists can be created simply.
Unordered (bulleted) list - Create a list by starting a line with `+`, `-`, or `*`. The text will automatically wrap to the proper indentation level. - Sub-lists are made by indenting 2 spaces before the marker. * Marker character change forces new list start: + Sub-sub list item 1 You can also indent an item under the sub-list + Sub-sub list item 2 + Sub-sub list item 3 - Easy peasy! Ordered list 1. Start with a number followed by a period 2. Continue with another number followed by a period 2. Tip: The actual number doesn't matter -- it'll auto-correct the count! 4. So you can use sequential numbers... 1. ...or keep all the numbers as `1.`
This renders as:
Unordered (bulleted) list
- Create a list by starting a line with +, -, or *. The text will automatically wrap to the proper indentation level.
- Sub-lists are made by indenting 2 spaces before the marker.
- Marker character change forces new list start:
-
- Sub-sub list item 1 You can also indent an item under the sub-list
- Sub-sub list item 2
- Sub-sub list item 3
- Easy peasy!
Ordered list
- Start with a number followed by a period
- Continue with another number followed by a period
- Tip: The actual number doesn't matter -- it'll auto-correct the count!
- So you can use sequential numbers...
- ...or keep all the numbers as 1.
Hyperlinks
Web links can also be included in a description. The link target can be visible or hidden behind other text.
https://google.com will automatically become a link [Google Search](https://google.com) link with custom text [Google Search](https://google.com "hover for a custom tip") [Link references][ref_name] [ref_name]: https://google.com "The Google search page"
This renders as:
Google Search link with custom text Google SearchLink references
Tables
Markdown can also be used to create a table. A few simple characters are used to create a structure that looks somewhat like a table in plain text. When rendered it becomes a beautiful HTML table.
A few tricks:
- Headers are optional and appear as the first line above a row with dashes. They appear bold and centered when rendered.
- Use colons on the line of dashes to indicate the justification for a column -- a single colon on the right to right justify, one on the right and left to center.
- Columns don’t actually have to line up nicely, just make sure there are at least three dashes in each column divider and matching number of | dividing bars. However you can exclude the first and last bars if you’d like.
| Header | Columns | Above | the | Line | | ------ | :------------------: | ------ | ------: | ---: | | Row 1 | Colons on line above | | right | 1.23 | | Row 2 | make this column | | aligned | 2.3 | Row 3|center content||column|4
This renders as:
| Header | Columns | Above | the | Line |
| Row 1 | Colons on line above | right | 1.23 | |
| Row 2 | make this column | aligned | 2.3 | |
| Row 3 | center content | column | 4 |
Images
Markdown also allows you to include images by referencing them via a URL. Note that no additional formatting can be performed on the image, so it will be shown at its native size and orientation.
 Or an clickable image: [](https://tripleblind.com)
![https://tripleblind.com/wp-content/uploads/2020/05/[email protected]](https://tripleblind.com/wp-content/uploads/2020/05/TB-Logo-color@2x.png){kind=link}
This renders as:
And even an image 🔗with a link.
Full Example
Pulling it all together, here is a rich description done in markdown:
Half of the well-known CIFAR-10 dataset of color images. This dataset of small
(32x32) images can be used for image classifications problems. Traditionally
it is used as a single dataset for training, but here it has been chopped in
half to illustrate distributed training.
Images are in one of ten different classes:
| | Class | Sample
|---|--------------|--------
| | _airplane_ | 
| | _automobile_ | 
| | _bird_ | 
| | _cat_ | 
| | _horse_ | 
| | _deer_ | 
| | _dog_ | 
| | _frog_ | 
| | _ship_ | 
| | _truck_ | 
Classes are mutually exclusive -- a "truck" is never an "automobile".
More information on the Canadian Institute for Advanced Research (CIFAR) and
this dataset can be found at [Alex Krizhevsky's CIFAR page](https://www.cs.toronto.edu/~kriz/cifar.html).
This renders as:
