Update to 150K iteration model checkpoint

- Updated app.py to load net_g_150000.pth instead of net_g_20000.pth
- Replaced model weights with more recent 150K iteration checkpoint
- Should provide improved super-resolution quality from extended training

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -681,7 +681,7 @@ model = HAT(
 )
 
 # Load the fine-tuned weights
-checkpoint = torch.load('net_g_20000.pth', map_location=device)
+checkpoint = torch.load('net_g_150000.pth', map_location=device)
 if 'params_ema' in checkpoint:
     model.load_state_dict(checkpoint['params_ema'])
 elif 'params' in checkpoint:

net_g_20000.pth → net_g_150000.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94d74fbc11a23bec569dd31b9f10b1b1033fd839cadabbef155d2beab3a3ffeb
-size 170284129
+oid sha256:54b2eda431411bfecc50c542ec863267ed683128b7aa4163c109776c627bb721
+size 170285859