fix: Support variable kernel sizes in CNN layer generation

Training script was hardcoded to generate cnn_conv3x3_* calls regardless
of actual kernel size, causing shader validation errors when layer 1 used
5×5 kernel (100 weights) but called 3×3 function (expected 36).

Changes:
- train_cnn.py: Generate correct conv function based on kernel_sizes[i]
- cnn_conv5x5.wgsl: Add cnn_conv5x5_7to4 and cnn_conv5x5_7to1 variants
- Regenerate cnn_layer.wgsl with correct function calls for [3,5,3]
- Document kernel size→function mapping in HOWTO.md

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

1 files changed, 10 insertions, 1 deletions

diff --git a/doc/HOWTO.md b/doc/HOWTO.md
index 2c813f7..07d480b 100644
--- a/doc/HOWTO.md
+++ b/doc/HOWTO.md
@@ -86,12 +86,21 @@ make run_util_tests     # Utility tests
 
 ---
 
-## Training 
+## Training
 
 ```bash
 ./training/train_cnn.py --layers 3 --kernel_sizes 3,5,3 --epochs 10000 --batch_size 8 --input training/input/ --target training/output/ --checkpoint-every 1000
 ```
 
+Generate shaders from checkpoint:
+```bash
+./training/train_cnn.py --export-only training/checkpoints/checkpoint_epoch_7000.pth
+```
+
+**Note:** Kernel sizes must match shader functions:
+- 3×3 kernel → `cnn_conv3x3_7to4` (36 weights: 9 pos × 4 channels)
+- 5×5 kernel → `cnn_conv5x5_7to4` (100 weights: 25 pos × 4 channels)
+
 ---
 
 ## Timeline