fix: add --max-seq-len to patch and remove || true from Dockerfile

The upstream openai_server.py gained --max-seq-len after v5 was built.
Add it to the patch so every build gets explicit sequence length control.
Also add max_seq_len=args.max_seq_len to FasterQwen3TTS.from_pretrained()
so the argument is actually used.

Remove || true from git apply so patch failures fail the build instead
of silently producing a broken image.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Dockerfile

@@ -22,7 +22,7 @@ RUN git clone --depth 1 --branch ${FASTER_QWEN3_TTS_REF} \
 
 # Apply DGX Spark patches (non_streaming_mode=True, per-voice temperature/top_k/top_p)
 COPY patches/openai_server.patch /tmp/
-RUN cd /app && git apply /tmp/openai_server.patch || true
+RUN cd /app && git apply /tmp/openai_server.patch
 
 # Create virtual environment (Ubuntu 24.04 enforces PEP 668)
 ENV VIRTUAL_ENV=/opt/venv

patches/openai_server.patch

@@ -25,3 +25,17 @@ index 61047ea..2b0c8bb 100644
                  )
 
          audio_arrays, sr = await loop.run_in_executor(None, _generate)
+@@ -306,4 +309,5 @@ def _parse_args():
+     p.add_argument("--host", default="0.0.0.0", help="Bind host (default: 0.0.0.0)")
+     p.add_argument("--port", type=int, default=8000, help="Bind port (default: 8000)")
+     p.add_argument("--device", default="cuda", help="Torch device (default: cuda)")
++    p.add_argument("--max-seq-len", type=int, default=4096, help="Max sequence length for CUDA graph static cache (default: 4096)")
+     return p.parse_args()
+@@ -341,6 +349,7 @@ def main():
+     logger.info("Loading model %s on %s …", args.model, args.device)
+     tts_model = FasterQwen3TTS.from_pretrained(
+         args.model,
+         device=args.device,
+         dtype=torch.bfloat16,
++        max_seq_len=args.max_seq_len,
+     )