Merge branch 'dev/rebase-v0170' of https://github.com/vllm-project/vllm-omni into dev/rebase-v0170

vllm_omni/benchmarks/metrics/metrics.py

@@ -134,7 +134,7 @@ def calculate_metrics(
     input_requests: list[SampleRequest],
     outputs: list[RequestFuncOutput],
     dur_s: float,
-    tokenizer: PreTrainedTokenizerBase | None,
+    tokenizer: TokenizerLike | None,
     selected_percentiles: list[float],
     goodput_config_dict: dict[str, float],
     task_type,

vllm_omni/benchmarks/patch/patch.py

@@ -313,7 +313,7 @@ async def benchmark(
     base_url: str,
     model_id: str,
     model_name: str,
-    tokenizer: PreTrainedTokenizerBase | None,
+    tokenizer: TokenizerLike | None,
     input_requests: list[SampleRequest],
     logprobs: int | None,
     request_rate: float,