Merge branch 'main' into fix_ut

docs/usage/troubleshooting.md

@@ -38,7 +38,7 @@ If other strategies don't solve the problem, it's likely that the vLLM instance
 - `export VLLM_LOG_STATS_INTERVAL=1.` to get log statistics more frequently for tracking running queue, waiting queue and cache hit states.
 - `export CUDA_LAUNCH_BLOCKING=1` to identify which CUDA kernel is causing the problem.
 - `export NCCL_DEBUG=TRACE` to turn on more logging for NCCL.
-- `export VLLM_TRACE_FUNCTION=1` to record all function calls for inspection in the log files to tell which function crashes or hangs. Do not use this flag unless absolutely needed for debugging, it will cause significant delays in startup time.
+- `export VLLM_TRACE_FUNCTION=1` to record all function calls for inspection in the log files to tell which function crashes or hangs. (WARNING: This flag will slow down the token generation by **over 100x**. Do not use unless absolutely needed.)
 
 ## Breakpoints
 

vllm/model_executor/layers/rotary_embedding/mrope.py

@@ -4,7 +4,6 @@
 
 import numpy as np
 import torch
-from transformers import PretrainedConfig
 
 from vllm.triton_utils import tl, triton
 
@@ -376,39 +375,6 @@ class MRotaryEmbedding(RotaryEmbedding):
     ) -> tuple[torch.Tensor, torch.Tensor | None]:
         return self.forward_native(positions, query, key, offsets)
 
-    @classmethod
-    def get_input_positions(
-        cls,
-        input_tokens: list[int],
-        hf_config: PretrainedConfig,
-        image_grid_thw: list[list[int]] | torch.Tensor | None,
-        video_grid_thw: list[list[int]] | torch.Tensor | None,
-        second_per_grid_ts: list[float] | None,
-        context_len: int = 0,
-        seq_len: int | None = None,
-        audio_feature_lengths: torch.Tensor | None = None,
-        use_audio_in_video: bool = False,
-    ) -> tuple[list[list[int]], int]:
-        """Get mrope input positions and delta value."""
-
-        image_grid_thw = [] if image_grid_thw is None else image_grid_thw
-        video_grid_thw = [] if video_grid_thw is None else video_grid_thw
-        second_per_grid_ts = [] if second_per_grid_ts is None else second_per_grid_ts
-
-        llm_positions, mrope_position_delta = cls.get_input_positions_tensor(
-            input_tokens=input_tokens,
-            hf_config=hf_config,
-            image_grid_thw=image_grid_thw,
-            video_grid_thw=video_grid_thw,
-            second_per_grid_ts=second_per_grid_ts,
-            context_len=context_len,
-            seq_len=seq_len,
-            audio_feature_lengths=audio_feature_lengths,
-            use_audio_in_video=use_audio_in_video,
-        )
-
-        return llm_positions.tolist(), mrope_position_delta
-
     @staticmethod
     def get_next_input_positions(
         mrope_position_delta: int,