Deployed 09d4475 with MkDocs version: 1.0.4

Trainer/Checkpointing/index.html

@@ -546,11 +546,11 @@
                 
                   <h1>Checkpointing</h1>
                 
-                <p>Lightning can automate saving and loading checkpoints.</p>
+                <p>i Lightning can automate saving and loading checkpoints.</p>
 <hr />
 <h3 id="model-saving">Model saving</h3>
 <p>To enable checkpointing, define the checkpoint callback and give it to the trainer.</p>
-<pre><code class="python">from pytorch_lightning.utils.pt_callbacks import ModelCheckpoint
+<pre><code class="python">from pytorch_lightning.callbacks import ModelCheckpoint
 
 checkpoint_callback = ModelCheckpoint(
     filepath='/path/to/store/weights.ckpt',

Trainer/Training Loop/index.html

@@ -627,9 +627,13 @@ trainer = Trainer(enable_early_stop=True)
 
 <hr />
 <h4 id="gradient-clipping">Gradient Clipping</h4>
-<p>Use this to turn off early stopping and run training to the <a href="#force-training-for-min-or-max-epochs">max_epoch</a></p>
+<p>Gradient clipping may be enabled to avoid exploding gradients.
+Specifically, this will <a href="https://pytorch.org/docs/stable/nn.html#torch.nn.utils.clip_grad_norm_">clip the gradient norm computed over all model parameters <em>together</em></a>.</p>
 <pre><code class="python"># DEFAULT (ie: don't clip)
 trainer = Trainer(gradient_clip=0)
+
+# clip gradients with norm above 0.5
+trainer = Trainer(gradient_clip=0.5)
 </code></pre>
 
 <hr />