Removing instantiation in forward (#8580)

### Description

This is a minor fix for the previous PR on this class.

### Types of changes
<!--- Put an `x` in all the boxes that apply, and remove the not
applicable items -->
- [x] Non-breaking change (fix or new feature that would not break
existing functionality).
- [ ] Breaking change (fix or new feature that would cause existing
functionality to change).
- [ ] New tests added to cover the changes.
- [ ] Integration tests passed locally by running `./runtests.sh -f -u
--net --coverage`.
- [ ] Quick tests passed locally by running `./runtests.sh --quick
--unittests --disttests`.
- [ ] In-line docstrings updated.
- [ ] Documentation updated, tested `make html` command in the `docs/`
folder.

---------

Signed-off-by: Eric Kerfoot <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: ericspod

monai/networks/nets/diffusion_model_unet.py

@@ -34,7 +34,6 @@
 import math
 from collections.abc import Sequence
 from functools import reduce
-from typing import Optional
 
 import numpy as np
 import torch
@@ -2016,7 +2015,7 @@ def __init__(
 
         last_dim_flattened = int(reduce(lambda x, y: x * y, input_shape) * channels[-1])
 
-        self.out: Optional[nn.Module] = nn.Sequential(
+        self.out: nn.Module = nn.Sequential(
             nn.Linear(last_dim_flattened, 512), nn.ReLU(), nn.Dropout(0.1), nn.Linear(512, self.out_channels)
         )
 
@@ -2063,9 +2062,6 @@ def forward(
         h = h.reshape(h.shape[0], -1)
 
         # 5. out
-        self.out = nn.Sequential(
-            nn.Linear(h.shape[1], 512), nn.ReLU(), nn.Dropout(0.1), nn.Linear(512, self.out_channels)
-        )
         output: torch.Tensor = self.out(h)
 
         return output