Address automated review feedback

Signed-off-by: Viacheslav Klimkov <vklimkov@nvidia.com>

Author: vklimkov-nvidia

examples/tts/easymagpie_vllm_omni/easymagpie_vllm_omni/codec/packed.py

@@ -299,8 +299,6 @@ def _uniform_cuda(
             if state_indices is None or query_start_loc is None or has_initial is None:
                 raise RuntimeError("incomplete codec prefill metadata")
 
-        batch_size = state_indices.numel()
-        sequence_rows = inputs.shape[0] // batch_size
         joined = gather_packed_state_inputs(
             inputs,
             self.kv_cache[0],
@@ -446,8 +444,6 @@ def _uniform_cuda(
             if state_indices is None or query_start_loc is None or has_initial is None:
                 raise RuntimeError("incomplete codec prefill metadata")
 
-        batch_size = state_indices.numel()
-        sequence_rows = inputs.shape[0] // batch_size
         joined = gather_packed_state_inputs(
             inputs,
             self.kv_cache[0],

examples/tts/easymagpie_vllm_omni/easymagpie_vllm_omni/serving_adapter.py

@@ -108,7 +108,7 @@ def _prompt_len(self, speaker_id: str) -> int:
             self._prompt_len_cache[speaker_id] = plen
             return plen
 
-        def validate(self, request: "OpenAICreateSpeechRequest") -> str | None:
+        def validate(self, request: OpenAICreateSpeechRequest) -> str | None:
             if not request.input or not request.input.strip():
                 return "Input text cannot be empty"
             extra = request.extra_params
@@ -118,7 +118,7 @@ def validate(self, request: "OpenAICreateSpeechRequest") -> str | None:
 
         async def build(
             self,
-            request: "OpenAICreateSpeechRequest",
+            request: OpenAICreateSpeechRequest,
             sampling_params_list: list,
             has_inline_ref_audio: bool,
         ) -> "PreparedRequest":
@@ -137,7 +137,7 @@ async def build(
             }
             return PreparedRequest(prompt=prompt, tts_params={}, model_type=MODEL_TYPE)
 
-        def build_streaming_spec(self, request: "OpenAICreateSpeechRequest") -> EasyMagpieStreamingSpec:
+        def build_streaming_spec(self, request: OpenAICreateSpeechRequest) -> EasyMagpieStreamingSpec:
             """Build the speaker prefill and tokenizer metadata without complete text."""
             speaker_id = (request.voice or _DEFAULT_SPEAKER).strip()
             model_path = Path(self._model_path())

examples/tts/easymagpie_vllm_omni/easymagpie_vllm_omni/serving_stream.py

@@ -359,6 +359,7 @@ async def send_audio() -> int:
             await websocket.send_json({"type": "session.done", "total_sentences": 1})
             completed = True
         except (WebSocketDisconnect, asyncio.TimeoutError):
+            # Client disconnects and input timeouts are expected terminal conditions.
             pass
         except Exception as error:
             logger.exception("Incremental EasyMagpie generation failed for %s", request_id)

examples/tts/easymagpie_vllm_omni/easymagpie_vllm_omni/stage_processors.py

@@ -224,6 +224,7 @@ def _resolve_speech_delay(transfer_manager: Any) -> int:
             try:
                 model_config = getter()
             except Exception:
+                # Version-specific getters may fail before initialization; use the existing config fallback.
                 pass
 
     hf_config = getattr(model_config, "hf_config", None)

examples/tts/easymagpie_vllm_omni/scripts/benchmark_incremental_server.py

@@ -137,7 +137,7 @@ async def send_tokens() -> None:
                         elif event_type == "error":
                             raise RuntimeError(event.get("message", str(event)))
                         elif event_type == "session.done":
-                            await sender
+                            _ = await sender
                             break
                 finally:
                     if not sender.done():

examples/tts/easymagpie_vllm_omni/scripts/benchmark_model.py

@@ -435,6 +435,7 @@ async def run_one_request(
             try:
                 await gen.aclose()
             except Exception:
+                # Cleanup failure must not replace the request result already recorded above.
                 pass
     return meter.result
 

examples/tts/easymagpie_vllm_omni/vllm_plugin_easymagpie_omni/__init__.py

@@ -29,6 +29,7 @@ def register() -> None:
     try:
         AutoConfig.register(EasyMagpieCodecConfig.model_type, EasyMagpieCodecConfig)
     except ValueError:
+        # Plugin reloads may encounter the same model type already registered.
         pass
     MODELS_CONFIG_MAP.setdefault(_CODEC_ARCH, MambaModelConfig)