Disable AOTI compile to dodge spaces==0.49.3 NVML regression

HF's build pipeline force-installs spaces==0.49.3 today, which has a
regression in zero.torch.patching._move() that NVML-asserts during
worker_init before the AOTI compile body runs. No way to override the
pin from inside requirements.txt. Skipping the optimize_pipeline_ call
gets the space booting on FA3 + Phr00t v18 alone. Restore once HF
propagates spaces==0.50.0 to the build pipeline.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

app.py

@@ -123,7 +123,10 @@ pipe.transformer.set_attn_processor(QwenDoubleStreamAttnProcessorFA3())
 
 # --- Ahead-of-time compilation ---
 # Note: optimize_pipeline_ handles text encoder offloading internally to save memory during torch.export
-optimize_pipeline_(pipe, image=[Image.new("RGB", (1024, 1024)), Image.new("RGB", (1024, 1024))], prompt="prompt")
+# DISABLED 2026-05-12: HF build pipeline force-pins spaces==0.49.3 which has a regression in
+# zero.torch.patching._move() — NVML assert during worker_init kills AOTI compile at startup.
+# Restore once HF bumps the pipeline to spaces==0.50.0+.
+# optimize_pipeline_(pipe, image=[Image.new("RGB", (1024, 1024)), Image.new("RGB", (1024, 1024))], prompt="prompt")
 
 # --- UI Constants and Helpers ---
 MAX_SEED = np.iinfo(np.int32).max

requirements.txt

@@ -1,6 +1,5 @@
 git+https://github.com/huggingface/diffusers.git
 
-spaces==0.50.0
 transformers
 accelerate
 safetensors