Switch assistant to gemma-4-12B (8-bit) + expose system prompt

fix: add torchvision for Qwen3.5-VL video processor backend

feat: upload generated images to R2 with prompt/params + requester uid in metadata

Fix RUNTIME_ERROR (kernels/transformers import) and add NoobAI-XL model selector

Update requirements.txt

Performance enhancements (#2)

Update requirements.txt

Update requirements.txt

Create requirements.txt