Fabric docs feedback 2/n (#16480)

docs/source-pytorch/fabric/guide/lightning_module.rst

@@ -100,10 +100,10 @@ It is up to you to call everything at the right place.
     model.train()
     for epoch in range(num_epochs):
         for i, batch in enumerate(dataloader):
+            optimizer.zero_grad()
             loss = model.training_step(batch, i)
             fabric.backward(loss)
             optimizer.step()
-            optimizer.zero_grad()
 
             # Control when hooks are called
             if condition:

examples/fabric/image_classifier/train_fabric.py

@@ -77,7 +77,6 @@ def run(hparams):
     # by the command line. See all options: `lightning run model --help`
     fabric = Fabric()
 
-    fabric.hparams = hparams
     seed_everything(hparams.seed)  # instead of torch.manual_seed(...)
 
     transform = T.Compose([T.ToTensor(), T.Normalize((0.1307,), (0.3081,))])