mirror of
https://github.com/wassname/pytorch-lightning.git
synced 2026-09-10 12:21:57 +08:00
Deployed 4b97319 with MkDocs version: 1.0.4
This commit is contained in:
+6
-4
@@ -713,8 +713,8 @@
|
||||
<h1>Home</h1>
|
||||
|
||||
<h6 id="new-project-quick-start">New project Quick Start</h6>
|
||||
<p>To start a new project you define two files, a LightningModule and a Trainer file. </p>
|
||||
<p>Here's an example of how you would use Lightning for research. </p>
|
||||
<p>To start a new project define two files, a LightningModule and a Trainer file. <br />
|
||||
To illustrate Lightning power and simplicity, here's an example of a typical research flow. </p>
|
||||
<h6 id="case-1-bert">Case 1: BERT</h6>
|
||||
<p>Let's say you're working on something like BERT but want to try different ways of training or even different networks.<br />
|
||||
You would define a single LightningModule and use flags to switch between your different ideas. </p>
|
||||
@@ -790,8 +790,10 @@ You would define a single LightningModule and use flags to switch between your d
|
||||
</pre></div>
|
||||
</td></tr></table>
|
||||
|
||||
<p>Notice that without writing any GPU or 16-bit specific code, your models gain that capability by<br />
|
||||
using Lightning. </p>
|
||||
<p>Notice a few things about this flow: <br />
|
||||
1. You're writing pure PyTorch... no unnecessary abstractions or new libraries to learn. <br />
|
||||
2. You get free GPU and 16-bit support without writing any of that code in your model. <br />
|
||||
3. You also get all of the capabilities below (without coding or testing yourself). </p>
|
||||
<hr />
|
||||
<h6 id="templates">Templates</h6>
|
||||
<ol>
|
||||
|
||||
File diff suppressed because one or more lines are too long
Binary file not shown.
Reference in New Issue
Block a user