perf: reduce container sizes and build times

- miku-stt: switch PyTorch CUDA -> CPU-only (~2.5 GB savings)
  - Silero VAD already runs on CPU via ONNX (onnx=True), CUDA PyTorch was waste
  - faster-whisper/CTranslate2 uses CUDA directly, no PyTorch GPU needed
  - torch+torchaudio layer: 3.3 GB -> 796 MB; total image 9+ GB -> 6.83 GB
  - Tested: Silero VAD loads (ONNX), Whisper loads on cuda, server ready

- llama-swap-rocm: add root .dockerignore to fix 31 GB build context
  - Dockerfile clones all sources from git, never COPYs from context
  - 19 GB of GGUF model files were being transferred on every build
  - Now excludes everything (*), near-zero context transfer

- anime-face-detector: add .dockerignore to exclude accumulated outputs
  - api/outputs/ (56 accumulated detection files) no longer baked into image
  - api/__pycache__/ and images/ also excluded

- .gitignore: remove .dockerignore exclusion so these files are tracked

stt-realtime/requirements-gpu-torch.txt

@@ -1,5 +1,7 @@
-# PyTorch with CUDA 12.8 support
-# Updated per RealtimeSTT PR #295 for better performance
-torch==2.7.1+cu128
-torchaudio==2.7.1+cu128
---index-url https://download.pytorch.org/whl/cu128
+# PyTorch CPU-only (used solely for Silero VAD which runs on CPU)
+# Silero VAD's OnnxWrapper uses torch tensors internally but does not need GPU.
+# Faster-Whisper/CTranslate2 handles GPU transcription via CUDA directly.
+# torchaudio is required by silero-vad's utils_vad.py top-level import.
+torch==2.7.1+cpu
+torchaudio==2.7.1+cpu
+--index-url https://download.pytorch.org/whl/cpu