bug still in NeRF-W latent appearance in coarse model

the appearance code in coarse model will still be None because params are not assigned in train.py

to fix this bug, we need to add param for coarse model in train.py and use appearence code for coarse model in rendering.py line 255.
@kwea123

Start by inspecting train.py to trace how parameters are assigned to the coarse model, then read rendering.py around line 255. Verify how the coarse model receives and uses the appearance code. Done means the coarse model's appearance code is no longer None and is used during rendering.

Written by the indexing model from the issue text.