fix(deps): add torchaudio + pin transformers<5

DiffSynth 2.0.7 imports:
- torchaudio (blanket import in diffsynth.core.data.operators)
- SiglipVisionTransformer from transformers.models.siglip (renamed/removed in
transformers 5.x; works fine in 4.45 - 4.57)

Without these pins, both local venv setup and the HF Space build pulled
transformers 5.x via DiffSynth's own pyproject (which lists 'transformers'
unbounded), and the import chain broke at app boot.

requirements.txt

@@ -4,8 +4,10 @@ gradio==5.50.0
 # Locally, install it ad-hoc if you want to test the @spaces.GPU decorator path.
 diffsynth @ git+https://github.com/modelscope/DiffSynth-Studio.git
 torch>=2.4
+torchaudio>=2.4   # DiffSynth blanket-imports torchaudio in diffsynth.core.data.operators
 safetensors>=0.4.5
 huggingface-hub>=0.27
+transformers>=4.45,<5.0  # DiffSynth 2.0.7 imports SiglipVisionTransformer (removed in transformers 5.x)
 
 # ControlNet preprocessors
 controlnet-aux>=0.0.9