biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[VLM] Remove BaseProcessingInfo.get_mm_max_tokens_per_item (#16408)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-04-11 00:06:58 +08:00
committed by GitHub
parent 7678fcd5b6
commit 83b824c8b4
39 changed files with 104 additions and 677 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
vllm/model_executor/models/qwen2_audio.py
View File

@@ -109,17 +109,6 @@ class Qwen2AudioProcessingInfo(BaseProcessingInfo):
def get_supported_mm_limits(self) -> Mapping[str, Optional[int]]:
return {"audio": None}
def get_mm_max_tokens_per_item(
self,
seq_len: int,
mm_counts: Mapping[str, int],
) -> Mapping[str, int]:
hf_config = self.get_hf_config()
max_source_positions = hf_config.audio_config.max_source_positions
max_output_lengths = (max_source_positions - 2) // 2 + 1
return {"audio": max_output_lengths}
class Qwen2AudioDummyInputsBuilder(
BaseDummyInputsBuilder[Qwen2AudioProcessingInfo]):