fix(cnn_v3/training): rebuild optimizer before loading state on resume past FiLM warmupHEADmain

When resuming a checkpoint saved after the FiLM warmup phase, the optimizer
was created with frozen (fewer) param groups, causing a size mismatch when
loading the saved optimizer state. Fix: detect ckpt['epoch'] >= film_warmup,
unfreeze FiLM MLP, and rebuild the optimizer before loading its state dict.

handoff(Gemini): train_cnn_v3.py --resume now works past epoch 1500.

0 files changed, 0 insertions, 0 deletions