fix(backend): release memory when unloading MLX models

Unloading a TTS/Whisper/LLM model on the MLX backend only dropped the
Python reference (`del self.model`). MLX keeps freed array buffers in
its own allocator pool for reuse instead of returning them to the OS,
so the process's memory footprint never actually shrank after unload
on Apple Silicon (the default backend there) until the process exited.

Add empty_mlx_cache() (backend/backends/base.py), wrapping
mx.clear_cache(), and call it from the three MLX unload_model()
implementations: MLXTTSBackend, MLXSTTBackend, MLXQwenLLMBackend.

Separately, the voice-clone prompt cache (backend/utils/cache.py) is a
process-lifetime dict populated by create_voice_prompt() across every
TTS engine, but nothing ever cleared it on model unload — only the
unrelated /tasks/clear-cache endpoint touched it. Add
clear_voice_prompt_memory_cache() (memory only, disk cache untouched
so a later generation still reloads the prompt instead of recomputing
it) and wire it into every TTS unload path (services/tts.py and the
qwen_custom_voice / generic branches of unload_model_by_config).
Whisper and the LLM backends never produce voice prompts, so their
unload paths are left alone.

Testing:
- New unit tests: backend/tests/test_mlx_unload_clears_cache.py,
  backend/tests/test_voice_prompt_cache_unload.py (8 tests, all pass).
- Verified end-to-end on Apple Silicon against real cached models
  (Qwen TTS 1.7B, Whisper Turbo, Qwen3 0.6B): loaded each via the
  running app, unloaded via the real /models/{name}/unload endpoint,
  and confirmed via mx.get_cache_memory()/get_active_memory() that the
  MLX allocator's cache drops to 0 on every cycle. Ran a real
  voice-clone generation end to end and confirmed the in-memory prompt
  cache goes from 1 entry to 0 on unload while the on-disk .prompt
  file is left intact.

backend/utils/cache.py

@@ -93,6 +93,19 @@ def cache_voice_prompt(
     torch.save(voice_prompt, cache_file)
 
 
+def clear_voice_prompt_memory_cache() -> None:
+    """
+    Drop the in-memory voice prompt cache without touching the disk cache.
+
+    Backends call this when a TTS model unloads: the cached prompts (tensors
+    or device-backed dicts produced by that model) would otherwise keep
+    referencing memory forever, since nothing else ever clears this
+    process-lifetime dict. The disk cache is left alone, so the next
+    generation just reloads the prompt from disk instead of recomputing it.
+    """
+    _memory_cache.clear()
+
+
 def clear_voice_prompt_cache() -> int:
     """
     Clear all voice prompt caches (memory and disk).